Recent Advances in Wearable Sensors for Health

IEEE SENSORS JOURNAL

1

Recent Advances in Wearable Sensors for Health Monitoring Mary M. Rodgers, Vinay M. Pai, and Richard S. Conroy

1 2 3 4 5 6 7 8 9

Abstract— Wearable sensor technology continues to advance and provide significant opportunities for improving personalized healthcare. In recent years, advances in flexible electronics, smart materials, and low-power computing and networking have reduced barriers to technology accessibility, integration, and cost, unleashing the potential for ubiquitous monitoring. This paper discusses recent advances in wearable sensors and systems that monitor movement, physiology, and environment, with a focus on applications for Parkinson’s disease, stroke, and head and neck injuries.

10

14

I. I NTRODUCTION

12

15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35

A

IE E Pr E oo f

13

Index Terms— Wearable sensors, biomedical and environmental monitoring, sensor systems, accelerometers, patient monitoring.

11

CCORDING to a May 2013 report on disruptive technologies by McKinsey Global Institute [1], the top four technologies likely to have a significant potential economic impact by 2025 are: 1) mobile internet, 2) automation of knowledge work, 3) the internet of things and 4) cloud computing. These disruptive technologies also form the basis for ubiquitous healthcare. Ubiquitous healthcare (UHC) is currently understood to encompass healthcare services that are available to everyone, independent of time and location. Systems that can fulfill the promise of delivering healthcare services at any time and any location will have significant implications for the treatment of chronic disease conditions as well as maintaining and encouraging healthy and independent living. Ubiquitous healthcare systems take advantage of a large number of hardware and software components, including Wireless Body Area Networks (WBANs), mobile devices and wireless cloud services, in order to achieve pervasive delivery. As outlined by Ogunduyile, et al. [2], a ubiquitous healthcare system must: 1) provide accessibility to several available services from an healthcare provider, 2) be flexible, Manuscript received May 30, 2014; revised August 14, 2014; accepted August 25, 2014. The associate editor coordinating the review of this paper and approving it for publication was Prof. Zheng Cui. M. M. Rodgers is with the National Institute of Biomedical Imaging and Bioengineering, National Institutes of Health, Bethesda, MD 20892 USA, and also with the Department of Physical Therapy and Rehabilitation Science, University of Maryland School of Medicine, Baltimore, MD 21201 USA (e-mail: [email protected]). V. M. Pai and R. S. Conroy are with the National Institute of Biomedical Imaging and Bioengineering, National Institutes of Health, Bethesda, MD 20892 USA (e-mail: [email protected]; [email protected]). Color versions of one or more of the figures in this paper are available online at http://ieeexplore.ieee.org. Digital Object Identifier 10.1109/JSEN.2014.2357257

Fig. 1. The “big picture” view of wearable sensors and their role in improving healthcare. Note the dual role of the patient: a) the sensors user and b) the decision-maker regarding the health and wellness personnel with whom she is willing to share the data obtained by these sensors.

3) provide security in information exchange, 4) enable remote health data acquisition, 5) provide personalized service, and 6) develop automatic decision making and response for diseased or healthy situations. Figure 1 illustrates that a systems approach is needed to integrate sensors with safe, secure and timely collection, dissemination and interpretation of data related to health status. It also highlights that the role of user and decision-maker may or may not overlap. Wearable sensors should be physically and technologically flexible to enable the monitoring of subjects in their natural environment. They have the potential to provide a rich stream of information that can transform the practice of medicine. Personal monitoring technologies have exploded over the past five years, with Google GlassTM , FitBitTM and The Nike+ FuelBandTM representative of the movement, and part of the bigger move towards an “internet of things”. As sensors become smarter and more ubiquitous, they will enable more comprehensive monitoring. The richness of the collected datasets should lead to better understanding of wellness and disease processes, ultimately resulting in better treatments and health outcomes. While smart glasses, fitness trackers and biometric wristbands represent the cusp of the consumer market, biomedical research is also taking advantage of wearable sensors, displays and processors for studying human and animal subjects in their natural environment. The goal of much of the current technology development, as shown in Figure 1, is to make the devices as seamless, non-obtrusive and close to the

U.S. Government work not protected by U.S. copyright.

36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62

2

65 66 67 68 69 70 71 72 73 74 75 76 77

78

79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118

“wear and forget” ideal as possible in order to achieve ubiquitous healthcare. This paper discusses advances in non-invasive sensors that monitor activity, physiologic function, and the environment and describes three clinical use cases. Although this paper focuses more on the technical capabilities of these devices, many researchers are also tackling the human factors involved, including how to reduce the barriers to the meaningful use of devices, minimizing physical discomfort for long-term monitoring, and addressing social stigma associated with visible monitoring of health. This paper is not designed to be a comprehensive review of wearable sensors, but describes a selection of recent technology advances and applications that highlight the broad potential of these systems to improve healthy and independent living. II. W EARABLE S ENSORS There have been several successful cases where technologies have moved out of the clinic to monitor patients going about their day-to-day life over extended periods. Perhaps the most notable of these is the ECG Holter monitor for detecting arrhythmias [3]. Wearable sensor systems are progressively becoming less obtrusive and more powerful, permitting monitoring of patients for longer periods of time in their normal environment. Current commercially available systems are compact, enclosed in durable packaging, and utilize either portable local storage or low-power radios to transmit data to remote servers [3], [4]. The development and refinement of novel fabrication techniques, sustainable power sources, inexpensive storage capacity and more efficient communication strategies are critical to continue this trend towards “wear and forget”. Sensors are primarily used to monitor three types of signals: activity, physiological and environmental. Data from these sensors can be collected, analyzed and made available to the wearers, caregivers, or healthcare professionals with the goal of improving the management and delivery of care, engaging patients and encouraging independent living. As shown in Figure 1, in addition to passive monitoring, interfacing with these sensors through local input and communication networks can be beneficial for engaging the wearer and may significantly impact adoption. For local input, flexible multitouch sensors have been developed which can be cut to any desired shape [5] while for readout, a range of technologies from organic light emitting diodes (OLED) devices [6] to electrochromic displays [7] and thermochromic indicators [8] have been demonstrated in principle. The emergence of body sensor networks, personal area networks, as well as lowpower communication protocols including ZigBee, Z-Wave, and Bluetooth have simplified the networking of sensors and the collection of multi-parameter datasets to provide a more comprehensive view of the local environment of the wearer. Flexible sensors, no longer constrained to planar geometries, have the potential to be one of the key technologies in helping to realize ubiquitous healthcare. The development of elastomeric and electrically-conductive polymers, ultrathin inorganics and organic semiconductors have enabled

flexible, stretchable electronic systems that can conform to daily life [9]. It is the compatibility of these flexible sensors with daily life and the ease with which they interface with other information communication technologies that has driven the widespread experimentation and investigation of their use for healthcare. Using state of the art fabrication techniques, substrates and circuits approaching 1 µm in thickness, bending radii less than 10 µm and weighing less than 1 mg/cm2, electronic devices can potentially be truly imperceptible [10]. Reviews describing these advances in fabrication have been recently published [11]–[16]. Non-invasive flexible healthcare devices fall into two main categories: 1) electronic skins (e-skins) that adhere to the body surface and 2) clothing-based or accessory-based devices where proximity is sufficient. In addition to lightweight flexible electronics, rapid advances in material science have opened doors to other potential benefits including optically transparency [17], [18], self-healing devices, light detection and harvesting [19] and bioelectrochemically powered sensors [20]. Although demonstrated individually, many of these advances have yet to be integrated into a fully functional device that has been tested in a non-controlled human environment.

IE E Pr E oo f

63 64


A. Activity Monitors

The analysis of movement can provide many insights into well-being, rehabilitation and fitness. Non-contact devices such as pedometers have been widely available for many decades. The concept of 10,000 steps representing the activity energy expenditure to balance the average calorific intake has been developed and refined over the past three decades and embraced by several public health campaigns. However, it has been the development of low-cost inertial sensors utilizing micro-electromechanical systems (MEMS), and sophisticated software for accurately detecting steps that has resulted in a dramatic rise in the availability and use of the personal activity monitors. For instance, many personal electronic devices, including some smartphones, music players and electronic pedometers can track movement with some degree of sensitivity. The most accurate sensors under ideal circumstances, and calibrated for healthy adults, are typically accurate to + / 3% [21]. − The incorporation of multiple sensors including accelerometers, gyroscopes, goniometers, force sensors and pressure sensors can provide more detailed insight into movement characteristics such as gait [22], falls [23], tremor and dyskinesia [24]. A number of recent reviews have been published describing the technology involved in these devices, placement considerations, the measurements which they are capable of, and the validation of these measurements [25]–[27]. The intersection of flexible electronics and activity monitoring provides a rich area of research opportunity. The conversion of mechanical to electrical energy using flexible polymers may prove attractive for energy harvesting [28], though energy density, comfort and durability are challenges for creating fully self-sustained systems. Nanotechnology, in the form of fiber strain-gauges [29] and carbon nanotubes for

119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141

142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174

RODGERS et al.: RECENT ADVANCES IN WEARABLE SENSORS FOR HEALTH MONITORING

176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193

194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230

detecting strain [30] can be readily included into clothing and other conformal designs, and can be prepared in transparent forms [31]. However, capturing the full range of human motions in a stretchable, nonrestrictive design and seamless integration with other system components are challenges that need to be addressed. One area of great promise is the integration of diagnostic and therapeutic systems into theranostic devices. Recently a multifunctional wearable device has been developed that records muscle activity and is integrated with a controlled transdermal delivery system for releasing nanoparticles [32]. There are also many intriguing technologies being developed that may impact future device design. Spray-on sensors [33], and self-healing polymers [34] are two examples of advances that may address some of the challenges with existing approaches in the longer-term. However, in the shorter term, the focus will potentially remain on integration and durability of technologies, improved algorithms for motion detection and validation of results against other technologies. B. Physiological Monitors

Fig. 2. Skin-adhering wearable sensor. Photo shows components of a 1 × 2 cm2 array include transistors, an antenna, power coils, and temperature sensors. (Photo courtesy of John Rogers, University of Illinois at Urbana-Champaign).

patch-based wearable sensors can be uploaded to a computer or mobile device, providing a snapshot of the user’s health. The information can be used to guide performance or behavioral modifications in support of sports and fitness or health and wellness goals. Flexible, temporary transfer tattoo-based sensors, or “electronic skin” or “epidermal electronics”, show great promise for analyzing metabolites [37] and many other potential applications [38] (figure 2). The direct printing of multifunctional devices onto the skin to increase durability and mitigate some surface interface issues is very attractive [39], [40], though the natural skin exfoliation process still limits the lifetime of these devices to a couple of weeks. For chemical analysis, one of the long-term goals is to realize lab-on-a-chip approaches that can analyze bodily secretions, such as sweat and saliva. Real-time sweat analysis can provide information on pH, electrolytes and hydration, and there have been efforts to build flexible, textilebased systems [41], [42]. Electrochemical approaches are more amenable to flexible designs than more sensitive or multiplexed optically-based technologies, though the tradeoff between complexity of measurement and complexity of system implementation has yet to be fully explored. Biosensors that identify biological molecules or pathogenic organisms are of particular interest for studying complex processes. The ability to detect hormones, enzymes, or lipids would greatly assist in monitoring organ function, viral or bacterial infections, and metabolic disorders. Technologies which may open up new types of measurements in this area include ionogels and organic electrochemical transistors [43]. These approaches have yet to be integrated in fully functional devices deployed in the field, though there is an unmet need to be able to track biochemistry on a routine or continuous basis. Reliably detecting and alerting wearers and caregivers to abnormal physiological conditions with sufficiently high sensitivity and specificity will be critical to achieving wider spread adoption, and will be needed before acceptance of semiautomated or closed-loop systems. The challenge in developing a reliable, non-invasive blood glucose sensor illustrates the gap that can exist between demonstrating what is possible and translating that to a universally useful system for providing accurate and reproducible information [44].

IE E Pr E oo f

175

For many healthcare use cases, it is highly desirable to have sensors capable of directly monitoring the physiology of the wearer in real-time. These sensors can measure biological, chemical or physical phenomena to assess physiology when in contact with the skin. The technology challenge is how to maintain consistent contact for extended periods and under different conditions, while the healthcare challenges are how to achieve a high sensitivity and specificity for detecting abnormal events in real-time. Maintaining consistent contact with the body is a significant challenge when exposed to the varying conditions of daily life, and there are different strategies to trying to achieve this. Traditionally adhesive-based and elastic-band approaches have been effective in many circumstances, though epidermal electronics and temporary tattoo approaches are increasingly being investigated [35]. When a sensor is in long-term contact with the body, a number of physical and electrical measurements can be made, including heart rate, breathing rate, blood oxygen saturation, ballistocardiography, blood pressure, electroencephalography (EEG), electrocardiography (ECG), electromyography (EMG) and skin temperature [36]. For EEG, ECG and EMG capacitive sensors are typically used to measure biopotentials, while for vital sign measurements such as heart rate, respiration rate and blood pressure, optical detection techniques such as photoplethysmography or piezoelectric strain sensors are generally used. The commercial availability of patch-based wearable sensors represents a significant advancement in personal monitoring device design, functionality and wear time. Examples include HealthPatchTM and MetriaTM , which are patch-based wearable biometric sensors that adhere to the user’s skin and continuously gather physiological, lifestyle information, and other indicators for up to seven days. The devices contain multiple sensors that enable monitoring of key health indicators such as heart rate, breathing rate, skin temperature, posture, steps taken, activity and sleep patterns. Data from the

3

231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272

4

273


C. Environmental Monitors

302

III. M EDICAL U SE C ASES

275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300

303 304 305 306 307 308

309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327

IE E Pr E oo f

301

Environmental monitoring is critical both for adding context to activity and physiological measurements, as well as monitoring hazards. Wearable sensors that are able to detect exposure to contaminants such as explosives, viral DNA, radioactivity or high concentrations of toxic gases like carbon monoxide, and monitoring of pollutants such as heavy metals, allergens such as pollen, and environmental conditions such as intense ultraviolet light could significantly improve health and safety. Sensors can also be used to monitor and augment senses, such as tracking eye movements, enhancing somatosensory feedback, and filtering background noise. Environmental safety monitoring, particularly for personnel involved in high risk activities, has been actively pursued for several decades, though the advent of flexible and integrated electronic devices has dramatically expanded their capabilities. Multi-parameter sensors are now being developed for avalanche rescue [45], emergency responders [46], and space travel [47] and are likely to become more ubiquitous. The analysis of volatile organic compounds (VOCs) in exhaled breath samples can potentially provide valuable information regarding the progression of some illnesses [48], [49]. While most of the technology involved in volatile-compound detection methods is bulky and expensive, the development of micro- and nano-scale technology has dramatically increased the sampling capabilities of these sensors, with the potential that they can be used as part of a ubiquitous healthcare system. A number of more detailed review articles have been written recently [50].

274

There are a number of medical uses for wearable sensors that can significantly impact the management of chronic disease and health hazards. The following use cases demonstrate the potential power of wearable sensors for the management of Parkinson’s disease, post-stroke rehabilitation, and the detection/tracking of head and neck injuries. A. Parkinson’s Disease

diagnosed with PD. The patients carried out scripted activities in a randomized order; however the temporal resolution was limited to 1 minute. Salarian et al. [53] have used tri-axial gyroscopes to detect tremors on a per-second basis from subjects who were made to perform a scripted sequence of activities. Their algorithm yielded 99.5% sensitivity on tremor-only data and 94.2% specificity on tremor-free data. However, this algorithm was not capable of distinguishing between tremors and dyskinesia. Roy et al. [54] have developed a set of dynamic neural network (DNN) classifiers which are implemented based on decisions made through a rule-based controller within the Integrated Processing and Understanding of Signals (IPUS) framework [55]. Using a multi-window approach, signal features were calculated across multiple time and frequency windows of sensor data and provided as input to DNNs trained to detect the patient’s mobility and motor states. This data analysis approach was used in conjunction with a single sensor providing a combination of both surface electromyographic data and tri-axial accelerometer data located at distal portions of each symptomatic limb. The Timed Up and Go (TUG) test is a well-known clinical test of mobility and fall risk; longer TUG times have been shown to be indicators of increased risk of fall in patient populations with PD or stroke [56]. Weiss, et al. [57] used body-fixed accelerometers to enhance the utility of the TUG test when evaluating PD patients. Subjects wore a 3D-accelerometer on their lower back while performing the TUG test, and used a multi-channel data logger carried in a pouch on their belt to save the data for post-processing. Acceleration signals were recorded for two timed TUG trials, and a number of acceleration-derived parameters such as jerk and range of acceleration amplitude were calculated. Recently, Cancela, et al. [58] have evaluated the feasibility of a wearable system based on a wireless body area network to assess the gait in PD patients. For this purpose, they used the PERFORM platform, a telematic platform for remote PD monitoring developed by a European academic-industrial consortium over the last few years [59]. The PERFORM platform consists of a set of four tri-axial accelerometers, located on each limb of the patient to track the limb movement, and one accelerometer and gyroscope attached to the belt of the patient to measure body acceleration and angular rate. The sensors used the ZigBee protocol to transfer data to a data logger device on the patient’s belt. At the end of each day, the patient transferred the data to a home-based computer where it was automatically processed.

Parkinson’s disease (PD) is the second most common neurodegenerative disorder after Alzheimer’s disease. Its prevalence in industrialized countries is estimated at 0.3% of the entire population and about 1% in people older than 60 years [51]. While a number of wearable sensors are being used for patients with Parkinson’s disease (PD), the most significant challenge is combining the data from these sensors to generate useful knowledge and actionable information. Machine learning algorithms are typically used to analyze the complex and unpredictable characteristics of wearable sensor data in order to study tracking of movement disorders in PD patients. The overlap of voluntary activities of daily life with the variety of motions corresponding to movement disorders can make it difficult to resolve and monitor the motor function in PD and is driving the need for better algorithms. Keijsers et al. [52] have utilized static neural networks to detect dyskinesia from accelerometer sensors worn by patients

B. Stroke Management

Intensive long-term rehabilitation post-stroke is an important factor in ensuring motor function recovery. Tracking changes in motor function can be used as a feedback tool for guiding the rehabilitation process [60]. Uswatte, et al. [61], [62] have shown that accelerometer data can provide objective information about real-world arm activity in stroke survivors. In their study, 169 stroke survivors undergoing constraint-induced movement therapy wore an

328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374

375 376 377 378 379 380 381 382 383


385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427

428 429 430 431 432 433 434 435 436 437 438 439

accelerometer on both wrists for a period of three days. The results indicated good patient compliance and showed that the ratio of activity recorded on impaired and unimpaired arm using accelerometers could be used to gather clinicallyrelevant information about upper extremity motor status. Prajapati et al. [63] performed a similar study for the lower extremities, using two wireless accelerometers placed on each leg to monitor walking in stroke survivors. Results showed that the system could monitor the quantity, symmetry and major biomechanical characteristics of walking. Also, Patel et al. [64] used accelerometers placed on the arm to derive accurate estimates of upper extremity functional ability. The researchers used a small subset of tasks from the Wolf Functional Ability Scale (FAS) [65] to derive estimates of the total FAS score via analysis of the accelerometer data. As the tasks selected from the FAS closely resemble tasks performed during the performance of activities of daily living, such a system could be used for unobtrusively monitoring functional ability in the patients’ home environment. Austin, et al. [66] developed a novel method for analyzing in-home collected gait velocities and demonstrated how the methodology of monitoring the evolution in gait velocity over time can identify changes associated with adverse outcomes. This method is applicable for both detecting acute changes in gait function and tracking longer-term changes that occur more slowly over time. Huang et al. have developed a micro-sensor based extremity rehabilitation system to evaluate motor impairment [67]. This system, embedded in the fabric of the garments, includes a combination of inertial sensors: tri-axial accelerometers, magnetometers and gyroscopes to capture motion, and enable the reconstruction of 3-D movement by the stroke patients. They have shown that their system can automatically measure the clinical Active Range of Motion (AROM) scale [68]. Another example of home-based rehabilitation technology is the Stroke Rehabilitation Exerciser developed by Philips Research [69]. The Stroke Rehab Exerciser coaches the patient through a sequence of exercises for motor retraining, which are prescribed by the physical therapist and uploaded to a patient unit. A wireless inertial sensor system records the patient’s movements, analyzes the data for deviations from a personal movement target and provides feedback to the patient and the therapist.

Fig. 3. Novel flexible biosensor development using nanotubes: (a) vertically aligned gold nanowire electrodes and (b) nanostructured textile nanobiosensor. (from Rai et al. (2014) [75]).

and low average root mean square errors, relative to the reference wired devices. Duc, et al. [74] recently used a wearable inertial sensor system to measure head and thorax kinematics and assessed the cervical spine mobility from these measurements. The wearable system consisted of two inertial sensors, one on the forehead and the second on the thorax, which were linked to a lightweight data-logger worn at the waist. Each inertial sensor included a tri-axial gyroscope and a tri-axial accelerometer. By fusing the angular velocity and acceleration obtained from these sensors, the researchers were able to compute 3D cervical angles. These measurements were in excellent concordance with the reference systems. Preliminary evaluations on a limited set of patients with cervical disk disease have shown that the system could detect differences in the range of motion similar to the reference system. Traumatic brain injury (TBI) is a major public health problem affecting all age groups and is the leading cause of death in young adults. Concerns have been raised about the potential long-term effects of repeated concussion, particularly in young athletes and adults in professions associated with frequent head injury, such as the military or contact sports. Technologies such as Reebok’s ChecklightTM for tracking impact are commercially available, and typically use accelerometers and gyroscopes to measure linear and rotation acceleration and duration of impact. It is unclear whether this information is sufficient to indicate a concussion or severity of the concussion. Rai, et al [75] have developed a wireless helmet-based health-monitoring system that may provide more information in this regard. The system is a network of flexible sensors woven or printed into a skullcap worn under a helmet. Using carbon nanotube textile nanostructures (figure 3), they have incorporated pressure sensors and flexible gyroscopes to track intensity, direction and location of impact force, as well as measure rotational motion of the head and body balance, along with lateral head motion and body balance. The cap also includes a collection of textile-based, dry sensors that measure electrical activity in the brain, including signs that may indicate the onset of mild traumatic brain injury. These sensors detect loss of consciousness, drowsiness, dizziness,

IE E Pr E oo f

384

C. Head and Neck Injuries

Reduction in the range of motion of the cervical spine has been found to be a useful indicator of physical disability in neck pain [70] and a predictor of poor outcome after whiplash injury [71], [72]. A new generation of wireless orientation uniaxial accelerometer and a magnetometer, in each orthogonal axis. Jasiewicz, et al. [73] evaluated IC3 orientation sensors (Intersense, Bedford, MA, USA) and showed that these wireless orientation sensors perform as well as the standard non-invasive electromagnetic devices used to measure cervical motion. These self-contained, portable and relatively inexpensive wireless sensors had very high cross-correlations

5

440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479

6

481 482 483 484 485 486 487

488

489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535

fatigue, anxiety and sensitivity to light. Finally, the skullcap includes a sensor to detect pulse rate and blood oxygen level. The system utilizes ZigBee and Bluetooth wireless telemetry to transmit data from sensors to a remote server or monitor. The goal for this type of sensor system is to provide real-time evaluation of head trauma and rapidly triage cases for conventional neuro-imaging follow-up with magnetic resonance imaging (MRI) or computerized tomography (CT). IV. C ONCLUSION AND F UTURE D IRECTIONS Recent advances in flexible electronics show great promise for healthcare monitoring. A great deal of work has been accomplished toward the integration of wearable technologies and communication [76] as well as data analysis technologies so that the goal of remote monitoring individuals in the home and community settings can be achieved. When monitoring has been performed in the home, researchers and clinicians have integrated ambient sensors in the remote monitoring systems. However some challenges remain, including efficient energy harvesting, human-device interfacing and improving the quality and range of measurements. The integration of different power sources, sensors and processing and testing in a non-controlled human environment is essential to establishing confidence in the diagnostic capabilities of these systems and their ability to change outcomes. One of the key areas of opportunity is the development and optimization of techniques to measure other physiologic metrics with higher accuracy. To date, we do not have objective, non-invasive techniques for assessing pain, mental state such as attention span and nervousness, biochemical status such as hormone levels, immunologic status for tracking exposure to infectious agents, and interactions with humans or companion animals. In addition, there is a need to develop and optimize accurate, robust and continuous physiological monitoring within closed-loop systems. For example, many non-invasive blood glucose devices have been developed, though their accuracy is limited by many confounding factors such as hydration, temperature, metabolic modulators, and comorbidities. A recent review found that 40% of continuous monitors satisfy the American Diabetes Association precision criteria of + /− 10% and fewer than 20% achieved + /− 5% accuracy. While two-thirds of these monitors were not sensitive enough to detect hypoglycemia, results from the Diabetes Control and Complications Trial suggest their potential in reducing longterm HbA1c levels [77]. A second key area of opportunity is the integration of environmental sensing and feedback to the wearer, enhancing sensory-feedback in the form of retinal and cochlear implants, and monitoring environmental risks such as radiation, pathogens, chemicals and poor air quality, and environmental parameters such as ambient temperature, lightlevels, humidity, proximity and location. With the increasing availability and reduction in cost of low power communication protocols and pervasiveness of automatic identification and data capture technologies, health care data is likely to be increasingly tagged with environmental data to describe context and enhance understanding.

A third key area of opportunity is the development of automatic algorithms that can generate relevant clinical alerts based on changes in the physiological data. Visualization tools are also needed to guide doctors, caregivers, and wearers around abnormal events, and identify increased risks that may lead to adverse health outcomes. Research toward achieving remote monitoring of older adults and subjects undergoing clinical interventions will soon face the need for establishing business models to cover the costs and identify reimbursement mechanisms for the technology and its management. Building a solid evidence base for the effectiveness of these sensor systems and addressing costs and reimbursement problems will be essential to assure that wearable sensor systems deliver on their promise of improving the quality of care for older adults and subjects affected by chronic conditions.

551

R EFERENCES

552

[1] J. Manyika, M. Chui, J. Bughin, R. Dobbs, P. Bisson, and A. Marrs, Disruptive Technologies: Advances that Will Transform Life, Business, and the Global Economy. McKinsey Global Institute, May 2013. [2] O. O. Ogunduyile, K. Zuva, O. A. Randle, and T. Zuva, “Ubiquitous healthcare monitoring system using integrated triaxial accelerometer, SpO2 and location sensors,” Int. J. UbiComp, vol. 4, pp. 1–13, Sep. 2013. [3] P. Bonato, “Wearable sensors/systems and their impact on biomedical engineering,” IEEE Eng. Med. Biol. Mag., vol. 22, no. 3, pp. 18–20, May/Jun. 2003. [4] P. Bonato, “Wearable sensors and systems,” IEEE Eng. Med. Biol. Mag., vol. 29, no. 3, pp. 25–36, May/Jun. 2010. [5] S. Olberding, N.-W. Gong, J. Tiab, J. A. Paradiso, and J. Steimle, “A cuttable multi-touch sensor,” in Proc. 26th Annu. ACM Symp. User Interf. Softw. Technol., 2013, pp. 245–254. [6] R. Ma et al., “Wearable 4-in. QVGA full-color-video flexible AMOLEDs for rugged applications,” J. Soc. Inf. Display, vol. 18, no. 1, pp. 50–56, 2010. [7] C. Yan et al., “Stretchable and wearable electrochromic devices,” ACS Nano, vol. 8, no. 1, pp. 316–322, Dec. 2014. [8] C. Yu, Y. Zhang, D. Cheng, X. Li, Y. Huang, and J. A. Rogers, “All-elastomeric, strain-responsive thermochromic color indicators,” Small, vol. 10, no. 7, pp. 1266–1271, 2014. [9] J. A. Rogers, T. Someya, and Y. G. Huang, “Materials and mechanics for stretchable electronics,” Science, vol. 327, no. 5973, pp. 1603–1607, Mar. 2010. [10] M. Kaltenbrunner et al., “An ultra-lightweight design for imperceptible plastic electronics,” Nature, vol. 499, pp. 458–463, Jul. 2013. [11] D.-H. Kim, J. Xiao, J. Song, Y. Huang, and J. A. Rogers, “Stretchable, curvilinear electronics based on inorganic materials,” Adv. Mater., vol. 22, no. 19, pp. 2108–2124, May 2010. [12] S. Park, M. Vosguerichian, and Z. Bao, “A review of fabrication and applications of carbon nanotube film-based flexible electronics,” Nanoscale, vol. 5, no. 5, pp. 1727–1752, 2013. [13] J. S. Park, W.-J. Maeng, H.-S. Kim, and J.-S. Park, “Review of recent developments in amorphous oxide semiconductor thin-film transistor devices,” Thin Solid Films, vol. 520, no. 6, pp. 1679–1693, Jan. 2012. [14] D.-H. Kim, R. Ghaffari, N. Lu, and J. A. Rogers, “Flexible and stretchable electronics for biointegrated devices,” Annu. Rev. Biomed. Eng., vol. 14, pp. 113–128, Aug. 2012. [15] W. Zeng, L. Shu, Q. Li, S. Chen, F. Wang, and X.-M. Tao, “Fiber-based wearable electronics: A review of materials, fabrication, devices, and applications,” Adv. Mater., vol. 26, no. 31, pp. 5310–5336, Aug. 2014. [16] A. Nathan et al., “Flexible electronics: The next ubiquitous platform,” Proc. IEEE, vol. 100, pp. 1486–1517, May 2012. [17] S. Bae et al., “Roll-to-roll production of 30-inch graphene films for transparent electrodes,” Nature Nanotechnol., vol. 5, pp. 574–578, Jun. 2010. [18] L. Hu, H. S. Kim, J.-Y. Lee, P. Peumans, and Y. Cui, “Scalable coating and properties of transparent, flexible, silver nanowire electrodes,” ACS Nano, vol. 4, no. 5, pp. 2955–2963, Apr. 2010. [19] G. Konstantatos and E. H. Sargent, “Nanostructured materials for photon detection,” Nature Nanotechnol., vol. 5, pp. 391–400, Jun. 2010.

IE E Pr E oo f

480


536 537 538 539 540 541 542 543 544 545 546 547 548 549 550

553 554 555

AQ:1

556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604

AQ:2


606 607 608 609 610 611 612 613 614 615

AQ:3

616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675

AQ:4

676 677 678 679

[20] M. Zhou and S. Dong, “Bioelectrochemical interface engineering: Toward the fabrication of electrochemical biosensors, biofuel cells, and self-powered logic biosensors,” Accounts Chem. Res., vol. 44, no. 11, pp. 1232–1243, Aug. 2011. [21] M. Karabulut, S. E. Crouter, and D. R. Bassett, Jr., “Comparison of two waist-mounted and two ankle-mounted electronic pedometers,” Eur. J. Appl. Physiol., vol. 95, no. 4, pp. 335–343, Oct. 2005. [22] W. Tao, T. Liu, R. Zheng, and H. Feng, “Gait analysis using wearable sensors,” Sensors, vol. 12, no. 2, pp. 2255–2283, Feb. 2012. [23] S. Chaudhuri, H. Thompson, and G. Demiris, “Fall detection devices and their use with older adults: A systematic review,” J. Geriatric Phys. Therapy, Jan. 2014. [24] B. T. Cole, S. H. Roy, C. J. De Luca, and S. H. Nawab, “Dynamical learning and tracking of tremor and dyskinesia from wearable sensors,” IEEE Trans. Neural Syst. Rehabil. Eng., vol. 22, no. 5, pp. 982–991, Sep. 2014. [25] C.-C. Yang and Y.-L. Hsu, “A review of accelerometry-based wearable motion detectors for physical activity monitoring,” Sensors, vol. 10, no. 8, pp. 7772–7788, 2010. [26] D. R. Bassett, Jr., A. V. Rowlands, and S. G. Trost, “Calibration and validation of wearable monitors,” Med. Sci. Sports Exerc., vol. 44, pp. S32–S38, Jan. 2012. [27] N. F. Butte, U. Ekelund, and K. R. Westerterp, “Assessing physical activity using wearable monitors: Measures of physical activity,” Med. Sci. Sports Exerc., vol. 44, pp. S5–S12, Jan. 2012. [28] F.-R. Fan, Z.-Q. Tian, and Z. L. Wang, “Flexible triboelectric generator,” Nano Energy, vol. 1, no. 2, pp. 328–334, Mar. 2012. [29] C. Pang et al., “A flexible and highly sensitive strain-gauge sensor using reversible interlocking of nanofibres,” Nature Mater., vol. 11, pp. 795–801, Jul. 2012. [30] T. Yamada et al., “A stretchable carbon nanotube strain sensor for human-motion detection,” Nature Nanotechnol., vol. 6, pp. 296–301, Mar. 2011. [31] D. J. Lipomi et al., “Skin-like pressure and strain sensors based on transparent elastic films of carbon nanotubes,” Nature Nanotechnol., vol. 6, pp. 788–792, Oct. 2011. [32] D. Son et al., “Multifunctional wearable devices for diagnosis and therapy of movement disorders,” Nature Nanotechnol., vol. 9, pp. 397–404, Mar. 2014. [33] S. Schreml et al., “A sprayable luminescent pH sensor and its use for wound imaging in vivo,” Experim. Dermatol., vol. 21, no. 12, pp. 951–953, Dec. 2012. [34] B. C.-K. Tee, C. Wang, R. Allen, and Z. Bao, “An electrically and mechanically self-healing composite with pressure- and flexion-sensitive properties for electronic skin applications,” Nature Nanotechnol., vol. 7, pp. 825–832, Nov. 2012. [35] D.-H. Kim et al., “Epidermal electronics,” Science, vol. 333, no. 6044, pp. 838–843, Aug. 2011. [36] J. R. Windmiller and J. Wang, “Wearable electrochemical sensors and biosensors: A review,” Electroanalysis, vol. 25, no. 1, pp. 29–46, Jan. 2013. [37] J. R. Windmiller, A. J. Bandodkar, G. Valdés-Ramírez, S. Parkhomovsky, A. G. Martinez, and J. Wang, “Electrochemical sensing based on printable temporary transfer tattoos,” Chem. Commun., vol. 48, no. 54, pp. 6794–6796, May 2012. [38] S. Bauer, “Flexible electronics: Sophisticated skin,” Nature Mater., vol. 12, pp. 871–872, Sep. 2013. [39] W.-H. Yeo et al., “Multifunctional epidermal electronics printed directly onto the skin,” Adv. Mater., vol. 25, no. 20, pp. 2773–2778, May 2013. [40] S. Xu et al., “Soft microfluidic assemblies of sensors, circuits, and radios for the skin,” Science, vol. 344, no. 6179, pp. 70–74, Apr. 2014. [41] D. Diamond, S. Coyle, S. Scarmagnani, and J. Hayes, “Wireless sensor networks and chemo-/biosensing,” Chem. Rev., vol. 108, no. 2, pp. 652–679, Jan. 2008. [42] D. Morris, S. Coyle, Y. Wu, K. T. Lau, G. Wallace, and D. Diamond, “Bio-sensing textile based patch with integrated optical detection system for sweat monitoring,” Sens. Actuat. B, Chem., vol. 139, no. 1, pp. 231–236, May 2009. [43] K. J. Fraser et al., “Wearable electrochemical sensors for monitoring performance athletes,” Proc. SPIE, Organic Semicond. Sensors Bioelectron. IV, vol. 8118, p. 81180C-1–81180C-12, Sep. 2011. [44] S. K. Vashist, “Non-invasive glucose monitoring technology in diabetes management: A review,” Anal. Chim. Acta, vol. 750, pp. 16–27, Oct. 2012.

[45] F. Michahelles, P. Matter, A. Schmidt, and B. Schiele, “Applying wearable sensors to avalanche rescue,” Comput. Graph., vol. 27, no. 6, pp. 839–847, Dec. 2003. [46] D. Curone et al., “Smart garments for safety improvement of emergency/disaster operators,” in Proc. 29th Annu. Int. Conf. IEEE Eng. Med. Biol. Soc., vol. 2007. Aug. 2007, pp. 3962–3965. [47] C. W. Mundt et al., “A multiparameter wearable physiologic monitoring system for space and terrestrial applications,” IEEE Trans. Inf. Technol. Biomed., vol. 9, no. 3, pp. 382–391, Sep. 2005. [48] A. W. Boots, J. J. B. N. van Berkel, J. W. Dallinga, A. Smolinska, E. F. Wouters, and F. J. van Schooten, “The versatile use of exhaled volatile organic compounds in human health and disease,” J. Breath Res., vol. 6, no. 2, p. 027108, Jun. 2012. [49] B. Buszewski, M. K¸esy, T. Ligor, and A. Amann, “Human exhaled air analytics: Biomarkers of diseases,” Biomed. Chromatogr., vol. 21, no. 6, pp. 553–566, Jun. 2007. [50] G. Konvalina and H. Haick, “Sensors for breath testing: From nanomaterials to comprehensive disease detection,” Accounts Chem. Res., vol. 47, no. 1, pp. 66–76, Jan. 2014. [51] L. M. L. de Lau and M. M. B. Breteler, “Epidemiology of Parkinson’s disease,” Lancet Neurol., vol. 5, no. 6, pp. 525–535, Jun. 2006. [52] N. L. W. Keijsers, M. W. I. M. Horstink, and S. C. A. M. Gielen, “Ambulatory motor assessment in Parkinson’s disease,” Movement Disorders, vol. 21, no. 1, pp. 34–44, Jan. 2006. [53] A. Salarian, H. Russmann, C. Wider, P. R. Burkhard, F. J. Vingerhoets, and K. Aminian, “Quantification of tremor and bradykinesia in Parkinson’s disease using a novel ambulatory monitoring system,” IEEE Trans. Biomed. Eng., vol. 54, no. 2, pp. 313–322, Feb. 2007. [54] S. H. Roy, B. T. Cole, L. D. Gilmore, C. J. De Luca, and S. H. Nawab, “Resolving signal complexities for ambulatory monitoring of motor function in Parkinson’s disease,” presented at the Annu. Int. Conf. IEEE Eng. Med. Biol. Soc. (EMBC), Boston, MA, USA, Aug./Sep. 2011, pp. 4832–4835. [55] V. R. Lesser, S. H. Nawab, and F. I. Klassner, “IPUS: An architecture for the integrated processing and understanding of signals,” Artif. Intell., vol. 77, no. 1, pp. 129–171, Aug. 1995. [56] Y. Balash, C. Peretz, G. Leibovich, T. Herman, J. M. Hausdorff, and N. Giladi, “Falls in outpatients with Parkinson’s disease: Frequency, impact and identifying factors,” J. Neurol., vol. 252, no. 11, pp. 1310–1315, 2005. [57] A. Weiss et al., “Can an accelerometer enhance the utility of the timed up & go test when evaluating patients with Parkinson’s disease?” Med. Eng. Phys., vol. 32, no. 2, pp. 119–125, 2010. [58] J. Cancela, M. Pastorino, M. T. Arredondo, N. S. Konstantina, F. Villagra, and M. A. Pastor, “Feasibility study of a wearable system based on a wireless body area network for gait assessment in Parkinson’s disease patients,” Sensors, vol. 14, no. 3, pp. 4618–4633, 2014. [59] J. Cancela, M. Pastorino, M. T. Arredondo, and O. Hurtado, “A telehealth system for Parkinson’s disease remote monitoring. The PERFORM approach,” in Proc. 35th Annu. Int. Conf. IEEE Eng. Med. Biol. Soc. (EMBC), Jul. 2013, pp. 7492–7495. [60] S. Patel, H. Park, P. Bonato, L. Chan, and M. Rodgers, “A review of wearable sensors and systems with application in rehabilitation,” J. Neuroeng. Rehabil., vol. 9, p. 21, Apr. 2012. [61] G. Uswatte, W. L. Foo, H. Olmstead, K. Lopez, A. Holand, and L. B. Simms, “Ambulatory monitoring of arm movement using accelerometry: An objective measure of upper-extremity rehabilitation in persons with chronic stroke,” Archives Phys. Med. Rehabil., vol. 86, no. 7, pp. 1498–1501, Jul. 2005. [62] G. Uswatte, C. Giuliani, C. Winstein, A. Zeringue, L. Hobbs, and S. L. Wolf, “Validity of accelerometry for monitoring real-world arm activity in patients with subacute stroke: Evidence from the extremity constraint-induced therapy evaluation trial,” Archives Phys. Med. Rehabil., vol. 87, no. 10, pp. 1340–1345, Oct. 2006. [63] S. K. Prajapati, W. H. Gage, D. Brooks, S. E. Black, and W. E. McIlroy, “A novel approach to ambulatory monitoring: Investigation into the quantity and control of everyday walking in patients with subacute stroke,” Neurorehabil. Neural Repair, vol. 25, no. 1, pp. 6–14, Jan. 2011. [64] S. Patel et al., “A novel approach to monitor rehabilitation outcomes in stroke survivors using wearable technology,” Proc. IEEE, vol. 98, no. 3, pp. 450–461, Mar. 2010. [65] S. L. Wolf, J. P. McJunkin, M. L. Swanson, and P. S. Weiss, “Pilot normative database for the wolf motor function test,” Archives Phys. Med. Rehabil., vol. 87, no. 3, pp. 443–445, Mar. 2006.

IE E Pr E oo f

605

7

680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754

8

757 758 759 760 761 762 763 764 765 766 767

AQ:5

768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794

[66] D. Austin, T. L. Hayes, J. Kaye, N. Mattek, and M. Pavel, “Unobtrusive monitoring of the longitudinal evolution of in-home gait velocity data with applications to elder care,” in Proc. Annu. Int. Conf. IEEE Eng. Med. Biol. Soc., vol. 2011, Aug./Sep. 2011, pp. 6495–6498. [67] S. Huang et al., “Motor impairment evaluation for upper limb in stroke patients on the basis of a microsensor,” Int. J. Rehabil. Res., vol. 35, no. 2, pp. 161–169, 2012. [68] I. Günal, N. Köse, O. Erdogan, E. Göktürk, and S. Seber, “Normal range of motion of the joints of the upper extremity in male subjects, with special reference to side,” J. Bone Joint Surgery Amer., vol. 78, no. 9, pp. 1401–1404, Sep. 1996. [69] G. Lanfermann, J. te Vrugt, A. Timmermans, E. Bongers, N. Lambert, and V. van Acht, “Philips stroke rehabilitation exerciser,” presented at the Tech. Aids Rehabil. (TAR), Aachen, Germany, Jan. 2007. [70] P. T. Dall’Alba, M. M. Sterling, J. M. Treleaven, S. L. Edwards, and G. A. Jull, “Cervical range of motion discriminates between asymptomatic persons and those with whiplash,” Spine, vol. 26, no. 19, pp. 2090–2094, 2001. [71] M. Sterling, G. Jull, B. Vicenzino, J. Kenardy, and R. Darnell, “Development of motor system dysfunction following whiplash injury,” Pain, vol. 103, nos. 1–2, pp. 65–73, 2003. [72] M. Sterling, “A proposed new classification system for whiplash associated disorders—Implications for assessment and management,” Manual Therapy, vol. 9, no. 2, pp. 60–70, 2004. [73] J. M. Jasiewicz, J. Treleaven, P. Condie, and G. Jull, “Wireless orientation sensors: Their suitability to measure head movement for neck pain assessment,” Manual Therapy, vol. 12, no. 4, pp. 380–385, 2007. [74] C. Duc, P. Salvia, A. Lubansu, V. Feipel, and K. Aminian, “A wearable inertial system to assess the cervical spine mobility: Comparison with an optoelectronic-based motion capture evaluation,” Med. Eng. Phys., vol. 36, no. 1, pp. 49–56, 2014. [75] P. Rai, S. Oh, P. Shyamkumar, M. Ramasamy, R. E. Harbaugh, and V. K. Varadan, “Nano- bio- textile sensors with mobile wireless platform for wearable health monitoring of neurological and cardiovascular disorders,” J. Electrochem. Soc., vol. 161, no. 2, pp. B3116–B3150, 2014. [76] R. Want, “iPhone: Smarter than the average phone,” IEEE Pervasive Comput., vol. 9, no. 3, pp. 6–9, Jul./Sep. 2010. [77] S. K. Vashist, “Continuous glucose monitoring systems: A review,” Diagnostics, vol. 3, no. 4, pp. 385–412, 2013.

Mary M. Rodgers received the B.S. and M.S. degrees in physical therapy from the University of North Carolina, Chapel Hill, NC, USA, in 1976 and 1981, respectively, and the Ph.D. degree in biomechanics from Pennsylvania State University, State College, PA, USA, in 1985. She is currently the George R. Hepburn Dynasplint Professor and the Vice Chair of the Department of Physical Therapy and Rehabilitation Science, University of Maryland School of Medicine, Baltimore, MD, USA, and an Advisor with the National Institute of Biomedical Imaging and Bioengineering, North Bethesda, MD, USA.

795

Vinay M. Pai received the B.E. degree in mechanical engineering and the M.E. degree from Saurashtra University, Rajkot, India, in 1991, the M.S.M.E. degree from the University of Iowa, Iowa City, IA, USA, in 1993, and the Ph.D. degree in mechanical engineering from Florida State University, Tallahassee, FL, USA, in 1997. He currently directs the Biomedical Imaging Informatics Program at the National Institute of Biomedical Imaging and Bioengineering, North Bethesda, MD, USA.

807

IE E Pr E oo f

755 756


Richard S. Conroy received the B.Sc. degree in laser physics and optoelectronics, and the Ph.D. degree in physics from the University of St. Andrews, St. Andrews, U.K., in 1994 and 1998, respectively, the A.L.M. degree in biotechnology from the Harvard Extension School, Cambridge, MA, USA, in 2005, and the M.I.M. degree in international management from the University of Maryland University College, Adelphi, MD, USA, in 2013. He directs the NIBIB division of applied science and technology.

796 797 798 799 800 801 802 803 804 805 806

808 809 810 811 812 813 814 815 816

817 818 819 820 821 822 823 824 825 826 827

AUTHOR QUERIES = = = = =

Please Please Please Please Please

provide the location for ref. [1]. provide the issue no. for ref. [16]. provide the volume no. and page range for ref. [23]. confirm the volume no. for refs. [43], [46], and [66]. provide the page range for ref. [69].

IE E Pr E oo f

AQ:1 AQ:2 AQ:3 AQ:4 AQ:5


1

Recent Advances in Wearable Sensors for Health Monitoring Mary M. Rodgers, Vinay M. Pai, and Richard S. Conroy

1 2 3 4 5 6 7 8 9

Abstract— Wearable sensor technology continues to advance and provide significant opportunities for improving personalized healthcare. In recent years, advances in flexible electronics, smart materials, and low-power computing and networking have reduced barriers to technology accessibility, integration, and cost, unleashing the potential for ubiquitous monitoring. This paper discusses recent advances in wearable sensors and systems that monitor movement, physiology, and environment, with a focus on applications for Parkinson’s disease, stroke, and head and neck injuries.

10

14

I. I NTRODUCTION

12

15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35

A

IE E Pr E oo f

13

Index Terms— Wearable sensors, biomedical and environmental monitoring, sensor systems, accelerometers, patient monitoring.

11

CCORDING to a May 2013 report on disruptive technologies by McKinsey Global Institute [1], the top four technologies likely to have a significant potential economic impact by 2025 are: 1) mobile internet, 2) automation of knowledge work, 3) the internet of things and 4) cloud computing. These disruptive technologies also form the basis for ubiquitous healthcare. Ubiquitous healthcare (UHC) is currently understood to encompass healthcare services that are available to everyone, independent of time and location. Systems that can fulfill the promise of delivering healthcare services at any time and any location will have significant implications for the treatment of chronic disease conditions as well as maintaining and encouraging healthy and independent living. Ubiquitous healthcare systems take advantage of a large number of hardware and software components, including Wireless Body Area Networks (WBANs), mobile devices and wireless cloud services, in order to achieve pervasive delivery. As outlined by Ogunduyile, et al. [2], a ubiquitous healthcare system must: 1) provide accessibility to several available services from an healthcare provider, 2) be flexible, Manuscript received May 30, 2014; revised August 14, 2014; accepted August 25, 2014. The associate editor coordinating the review of this paper and approving it for publication was Prof. Zheng Cui. M. M. Rodgers is with the National Institute of Biomedical Imaging and Bioengineering, National Institutes of Health, Bethesda, MD 20892 USA, and also with the Department of Physical Therapy and Rehabilitation Science, University of Maryland School of Medicine, Baltimore, MD 21201 USA (e-mail: [email protected]). V. M. Pai and R. S. Conroy are with the National Institute of Biomedical Imaging and Bioengineering, National Institutes of Health, Bethesda, MD 20892 USA (e-mail: [email protected]; [email protected]). Color versions of one or more of the figures in this paper are available online at http://ieeexplore.ieee.org. Digital Object Identifier 10.1109/JSEN.2014.2357257

Fig. 1. The “big picture” view of wearable sensors and their role in improving healthcare. Note the dual role of the patient: a) the sensors user and b) the decision-maker regarding the health and wellness personnel with whom she is willing to share the data obtained by these sensors.

3) provide security in information exchange, 4) enable remote health data acquisition, 5) provide personalized service, and 6) develop automatic decision making and response for diseased or healthy situations. Figure 1 illustrates that a systems approach is needed to integrate sensors with safe, secure and timely collection, dissemination and interpretation of data related to health status. It also highlights that the role of user and decision-maker may or may not overlap. Wearable sensors should be physically and technologically flexible to enable the monitoring of subjects in their natural environment. They have the potential to provide a rich stream of information that can transform the practice of medicine. Personal monitoring technologies have exploded over the past five years, with Google GlassTM , FitBitTM and The Nike+ FuelBandTM representative of the movement, and part of the bigger move towards an “internet of things”. As sensors become smarter and more ubiquitous, they will enable more comprehensive monitoring. The richness of the collected datasets should lead to better understanding of wellness and disease processes, ultimately resulting in better treatments and health outcomes. While smart glasses, fitness trackers and biometric wristbands represent the cusp of the consumer market, biomedical research is also taking advantage of wearable sensors, displays and processors for studying human and animal subjects in their natural environment. The goal of much of the current technology development, as shown in Figure 1, is to make the devices as seamless, non-obtrusive and close to the

U.S. Government work not protected by U.S. copyright.

36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62

2

64 65 66 67 68 69 70 71 72 73 74 75 76 77

78

79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118

“wear and forget” ideal as possible in order to achieve ubiquitous healthcare. This paper discusses advances in non-invasive sensors that monitor activity, physiologic function, and the environment and describes three clinical use cases. Although this paper focuses more on the technical capabilities of these devices, many researchers are also tackling the human factors involved, including how to reduce the barriers to the meaningful use of devices, minimizing physical discomfort for long-term monitoring, and addressing social stigma associated with visible monitoring of health. This paper is not designed to be a comprehensive review of wearable sensors, but describes a selection of recent technology advances and applications that highlight the broad potential of these systems to improve healthy and independent living. II. W EARABLE S ENSORS There have been several successful cases where technologies have moved out of the clinic to monitor patients going about their day-to-day life over extended periods. Perhaps the most notable of these is the ECG Holter monitor for detecting arrhythmias [3]. Wearable sensor systems are progressively becoming less obtrusive and more powerful, permitting monitoring of patients for longer periods of time in their normal environment. Current commercially available systems are compact, enclosed in durable packaging, and utilize either portable local storage or low-power radios to transmit data to remote servers [3], [4]. The development and refinement of novel fabrication techniques, sustainable power sources, inexpensive storage capacity and more efficient communication strategies are critical to continue this trend towards “wear and forget”. Sensors are primarily used to monitor three types of signals: activity, physiological and environmental. Data from these sensors can be collected, analyzed and made available to the wearers, caregivers, or healthcare professionals with the goal of improving the management and delivery of care, engaging patients and encouraging independent living. As shown in Figure 1, in addition to passive monitoring, interfacing with these sensors through local input and communication networks can be beneficial for engaging the wearer and may significantly impact adoption. For local input, flexible multitouch sensors have been developed which can be cut to any desired shape [5] while for readout, a range of technologies from organic light emitting diodes (OLED) devices [6] to electrochromic displays [7] and thermochromic indicators [8] have been demonstrated in principle. The emergence of body sensor networks, personal area networks, as well as lowpower communication protocols including ZigBee, Z-Wave, and Bluetooth have simplified the networking of sensors and the collection of multi-parameter datasets to provide a more comprehensive view of the local environment of the wearer. Flexible sensors, no longer constrained to planar geometries, have the potential to be one of the key technologies in helping to realize ubiquitous healthcare. The development of elastomeric and electrically-conductive polymers, ultrathin inorganics and organic semiconductors have enabled

flexible, stretchable electronic systems that can conform to daily life [9]. It is the compatibility of these flexible sensors with daily life and the ease with which they interface with other information communication technologies that has driven the widespread experimentation and investigation of their use for healthcare. Using state of the art fabrication techniques, substrates and circuits approaching 1 µm in thickness, bending radii less than 10 µm and weighing less than 1 mg/cm2, electronic devices can potentially be truly imperceptible [10]. Reviews describing these advances in fabrication have been recently published [11]–[16]. Non-invasive flexible healthcare devices fall into two main categories: 1) electronic skins (e-skins) that adhere to the body surface and 2) clothing-based or accessory-based devices where proximity is sufficient. In addition to lightweight flexible electronics, rapid advances in material science have opened doors to other potential benefits including optically transparency [17], [18], self-healing devices, light detection and harvesting [19] and bioelectrochemically powered sensors [20]. Although demonstrated individually, many of these advances have yet to be integrated into a fully functional device that has been tested in a non-controlled human environment.

IE E Pr E oo f

63


A. Activity Monitors

The analysis of movement can provide many insights into well-being, rehabilitation and fitness. Non-contact devices such as pedometers have been widely available for many decades. The concept of 10,000 steps representing the activity energy expenditure to balance the average calorific intake has been developed and refined over the past three decades and embraced by several public health campaigns. However, it has been the development of low-cost inertial sensors utilizing micro-electromechanical systems (MEMS), and sophisticated software for accurately detecting steps that has resulted in a dramatic rise in the availability and use of the personal activity monitors. For instance, many personal electronic devices, including some smartphones, music players and electronic pedometers can track movement with some degree of sensitivity. The most accurate sensors under ideal circumstances, and calibrated for healthy adults, are typically accurate to + / 3% [21]. − The incorporation of multiple sensors including accelerometers, gyroscopes, goniometers, force sensors and pressure sensors can provide more detailed insight into movement characteristics such as gait [22], falls [23], tremor and dyskinesia [24]. A number of recent reviews have been published describing the technology involved in these devices, placement considerations, the measurements which they are capable of, and the validation of these measurements [25]–[27]. The intersection of flexible electronics and activity monitoring provides a rich area of research opportunity. The conversion of mechanical to electrical energy using flexible polymers may prove attractive for energy harvesting [28], though energy density, comfort and durability are challenges for creating fully self-sustained systems. Nanotechnology, in the form of fiber strain-gauges [29] and carbon nanotubes for

119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141

142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174


176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193

194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230

detecting strain [30] can be readily included into clothing and other conformal designs, and can be prepared in transparent forms [31]. However, capturing the full range of human motions in a stretchable, nonrestrictive design and seamless integration with other system components are challenges that need to be addressed. One area of great promise is the integration of diagnostic and therapeutic systems into theranostic devices. Recently a multifunctional wearable device has been developed that records muscle activity and is integrated with a controlled transdermal delivery system for releasing nanoparticles [32]. There are also many intriguing technologies being developed that may impact future device design. Spray-on sensors [33], and self-healing polymers [34] are two examples of advances that may address some of the challenges with existing approaches in the longer-term. However, in the shorter term, the focus will potentially remain on integration and durability of technologies, improved algorithms for motion detection and validation of results against other technologies. B. Physiological Monitors

Fig. 2. Skin-adhering wearable sensor. Photo shows components of a 1 × 2 cm2 array include transistors, an antenna, power coils, and temperature sensors. (Photo courtesy of John Rogers, University of Illinois at Urbana-Champaign).

patch-based wearable sensors can be uploaded to a computer or mobile device, providing a snapshot of the user’s health. The information can be used to guide performance or behavioral modifications in support of sports and fitness or health and wellness goals. Flexible, temporary transfer tattoo-based sensors, or “electronic skin” or “epidermal electronics”, show great promise for analyzing metabolites [37] and many other potential applications [38] (figure 2). The direct printing of multifunctional devices onto the skin to increase durability and mitigate some surface interface issues is very attractive [39], [40], though the natural skin exfoliation process still limits the lifetime of these devices to a couple of weeks. For chemical analysis, one of the long-term goals is to realize lab-on-a-chip approaches that can analyze bodily secretions, such as sweat and saliva. Real-time sweat analysis can provide information on pH, electrolytes and hydration, and there have been efforts to build flexible, textilebased systems [41], [42]. Electrochemical approaches are more amenable to flexible designs than more sensitive or multiplexed optically-based technologies, though the tradeoff between complexity of measurement and complexity of system implementation has yet to be fully explored. Biosensors that identify biological molecules or pathogenic organisms are of particular interest for studying complex processes. The ability to detect hormones, enzymes, or lipids would greatly assist in monitoring organ function, viral or bacterial infections, and metabolic disorders. Technologies which may open up new types of measurements in this area include ionogels and organic electrochemical transistors [43]. These approaches have yet to be integrated in fully functional devices deployed in the field, though there is an unmet need to be able to track biochemistry on a routine or continuous basis. Reliably detecting and alerting wearers and caregivers to abnormal physiological conditions with sufficiently high sensitivity and specificity will be critical to achieving wider spread adoption, and will be needed before acceptance of semiautomated or closed-loop systems. The challenge in developing a reliable, non-invasive blood glucose sensor illustrates the gap that can exist between demonstrating what is possible and translating that to a universally useful system for providing accurate and reproducible information [44].

IE E Pr E oo f

175

For many healthcare use cases, it is highly desirable to have sensors capable of directly monitoring the physiology of the wearer in real-time. These sensors can measure biological, chemical or physical phenomena to assess physiology when in contact with the skin. The technology challenge is how to maintain consistent contact for extended periods and under different conditions, while the healthcare challenges are how to achieve a high sensitivity and specificity for detecting abnormal events in real-time. Maintaining consistent contact with the body is a significant challenge when exposed to the varying conditions of daily life, and there are different strategies to trying to achieve this. Traditionally adhesive-based and elastic-band approaches have been effective in many circumstances, though epidermal electronics and temporary tattoo approaches are increasingly being investigated [35]. When a sensor is in long-term contact with the body, a number of physical and electrical measurements can be made, including heart rate, breathing rate, blood oxygen saturation, ballistocardiography, blood pressure, electroencephalography (EEG), electrocardiography (ECG), electromyography (EMG) and skin temperature [36]. For EEG, ECG and EMG capacitive sensors are typically used to measure biopotentials, while for vital sign measurements such as heart rate, respiration rate and blood pressure, optical detection techniques such as photoplethysmography or piezoelectric strain sensors are generally used. The commercial availability of patch-based wearable sensors represents a significant advancement in personal monitoring device design, functionality and wear time. Examples include HealthPatchTM and MetriaTM , which are patch-based wearable biometric sensors that adhere to the user’s skin and continuously gather physiological, lifestyle information, and other indicators for up to seven days. The devices contain multiple sensors that enable monitoring of key health indicators such as heart rate, breathing rate, skin temperature, posture, steps taken, activity and sleep patterns. Data from the

3

231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272

4

273


C. Environmental Monitors

302

III. M EDICAL U SE C ASES

275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300

303 304 305 306 307 308

309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327

IE E Pr E oo f

301

Environmental monitoring is critical both for adding context to activity and physiological measurements, as well as monitoring hazards. Wearable sensors that are able to detect exposure to contaminants such as explosives, viral DNA, radioactivity or high concentrations of toxic gases like carbon monoxide, and monitoring of pollutants such as heavy metals, allergens such as pollen, and environmental conditions such as intense ultraviolet light could significantly improve health and safety. Sensors can also be used to monitor and augment senses, such as tracking eye movements, enhancing somatosensory feedback, and filtering background noise. Environmental safety monitoring, particularly for personnel involved in high risk activities, has been actively pursued for several decades, though the advent of flexible and integrated electronic devices has dramatically expanded their capabilities. Multi-parameter sensors are now being developed for avalanche rescue [45], emergency responders [46], and space travel [47] and are likely to become more ubiquitous. The analysis of volatile organic compounds (VOCs) in exhaled breath samples can potentially provide valuable information regarding the progression of some illnesses [48], [49]. While most of the technology involved in volatile-compound detection methods is bulky and expensive, the development of micro- and nano-scale technology has dramatically increased the sampling capabilities of these sensors, with the potential that they can be used as part of a ubiquitous healthcare system. A number of more detailed review articles have been written recently [50].

274

There are a number of medical uses for wearable sensors that can significantly impact the management of chronic disease and health hazards. The following use cases demonstrate the potential power of wearable sensors for the management of Parkinson’s disease, post-stroke rehabilitation, and the detection/tracking of head and neck injuries. A. Parkinson’s Disease

diagnosed with PD. The patients carried out scripted activities in a randomized order; however the temporal resolution was limited to 1 minute. Salarian et al. [53] have used tri-axial gyroscopes to detect tremors on a per-second basis from subjects who were made to perform a scripted sequence of activities. Their algorithm yielded 99.5% sensitivity on tremor-only data and 94.2% specificity on tremor-free data. However, this algorithm was not capable of distinguishing between tremors and dyskinesia. Roy et al. [54] have developed a set of dynamic neural network (DNN) classifiers which are implemented based on decisions made through a rule-based controller within the Integrated Processing and Understanding of Signals (IPUS) framework [55]. Using a multi-window approach, signal features were calculated across multiple time and frequency windows of sensor data and provided as input to DNNs trained to detect the patient’s mobility and motor states. This data analysis approach was used in conjunction with a single sensor providing a combination of both surface electromyographic data and tri-axial accelerometer data located at distal portions of each symptomatic limb. The Timed Up and Go (TUG) test is a well-known clinical test of mobility and fall risk; longer TUG times have been shown to be indicators of increased risk of fall in patient populations with PD or stroke [56]. Weiss, et al. [57] used body-fixed accelerometers to enhance the utility of the TUG test when evaluating PD patients. Subjects wore a 3D-accelerometer on their lower back while performing the TUG test, and used a multi-channel data logger carried in a pouch on their belt to save the data for post-processing. Acceleration signals were recorded for two timed TUG trials, and a number of acceleration-derived parameters such as jerk and range of acceleration amplitude were calculated. Recently, Cancela, et al. [58] have evaluated the feasibility of a wearable system based on a wireless body area network to assess the gait in PD patients. For this purpose, they used the PERFORM platform, a telematic platform for remote PD monitoring developed by a European academic-industrial consortium over the last few years [59]. The PERFORM platform consists of a set of four tri-axial accelerometers, located on each limb of the patient to track the limb movement, and one accelerometer and gyroscope attached to the belt of the patient to measure body acceleration and angular rate. The sensors used the ZigBee protocol to transfer data to a data logger device on the patient’s belt. At the end of each day, the patient transferred the data to a home-based computer where it was automatically processed.

Parkinson’s disease (PD) is the second most common neurodegenerative disorder after Alzheimer’s disease. Its prevalence in industrialized countries is estimated at 0.3% of the entire population and about 1% in people older than 60 years [51]. While a number of wearable sensors are being used for patients with Parkinson’s disease (PD), the most significant challenge is combining the data from these sensors to generate useful knowledge and actionable information. Machine learning algorithms are typically used to analyze the complex and unpredictable characteristics of wearable sensor data in order to study tracking of movement disorders in PD patients. The overlap of voluntary activities of daily life with the variety of motions corresponding to movement disorders can make it difficult to resolve and monitor the motor function in PD and is driving the need for better algorithms. Keijsers et al. [52] have utilized static neural networks to detect dyskinesia from accelerometer sensors worn by patients

B. Stroke Management

Intensive long-term rehabilitation post-stroke is an important factor in ensuring motor function recovery. Tracking changes in motor function can be used as a feedback tool for guiding the rehabilitation process [60]. Uswatte, et al. [61], [62] have shown that accelerometer data can provide objective information about real-world arm activity in stroke survivors. In their study, 169 stroke survivors undergoing constraint-induced movement therapy wore an

328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374

375 376 377 378 379 380 381 382 383


386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427

428 429 430 431 432 433 434 435 436 437 438 439

accelerometer on both wrists for a period of three days. The results indicated good patient compliance and showed that the ratio of activity recorded on impaired and unimpaired arm using accelerometers could be used to gather clinicallyrelevant information about upper extremity motor status. Prajapati et al. [63] performed a similar study for the lower extremities, using two wireless accelerometers placed on each leg to monitor walking in stroke survivors. Results showed that the system could monitor the quantity, symmetry and major biomechanical characteristics of walking. Also, Patel et al. [64] used accelerometers placed on the arm to derive accurate estimates of upper extremity functional ability. The researchers used a small subset of tasks from the Wolf Functional Ability Scale (FAS) [65] to derive estimates of the total FAS score via analysis of the accelerometer data. As the tasks selected from the FAS closely resemble tasks performed during the performance of activities of daily living, such a system could be used for unobtrusively monitoring functional ability in the patients’ home environment. Austin, et al. [66] developed a novel method for analyzing in-home collected gait velocities and demonstrated how the methodology of monitoring the evolution in gait velocity over time can identify changes associated with adverse outcomes. This method is applicable for both detecting acute changes in gait function and tracking longer-term changes that occur more slowly over time. Huang et al. have developed a micro-sensor based extremity rehabilitation system to evaluate motor impairment [67]. This system, embedded in the fabric of the garments, includes a combination of inertial sensors: tri-axial accelerometers, magnetometers and gyroscopes to capture motion, and enable the reconstruction of 3-D movement by the stroke patients. They have shown that their system can automatically measure the clinical Active Range of Motion (AROM) scale [68]. Another example of home-based rehabilitation technology is the Stroke Rehabilitation Exerciser developed by Philips Research [69]. The Stroke Rehab Exerciser coaches the patient through a sequence of exercises for motor retraining, which are prescribed by the physical therapist and uploaded to a patient unit. A wireless inertial sensor system records the patient’s movements, analyzes the data for deviations from a personal movement target and provides feedback to the patient and the therapist.

Fig. 3. Novel flexible biosensor development using nanotubes: (a) vertically aligned gold nanowire electrodes and (b) nanostructured textile nanobiosensor. (from Rai et al. (2014) [75]).

and low average root mean square errors, relative to the reference wired devices. Duc, et al. [74] recently used a wearable inertial sensor system to measure head and thorax kinematics and assessed the cervical spine mobility from these measurements. The wearable system consisted of two inertial sensors, one on the forehead and the second on the thorax, which were linked to a lightweight data-logger worn at the waist. Each inertial sensor included a tri-axial gyroscope and a tri-axial accelerometer. By fusing the angular velocity and acceleration obtained from these sensors, the researchers were able to compute 3D cervical angles. These measurements were in excellent concordance with the reference systems. Preliminary evaluations on a limited set of patients with cervical disk disease have shown that the system could detect differences in the range of motion similar to the reference system. Traumatic brain injury (TBI) is a major public health problem affecting all age groups and is the leading cause of death in young adults. Concerns have been raised about the potential long-term effects of repeated concussion, particularly in young athletes and adults in professions associated with frequent head injury, such as the military or contact sports. Technologies such as Reebok’s ChecklightTM for tracking impact are commercially available, and typically use accelerometers and gyroscopes to measure linear and rotation acceleration and duration of impact. It is unclear whether this information is sufficient to indicate a concussion or severity of the concussion. Rai, et al [75] have developed a wireless helmet-based health-monitoring system that may provide more information in this regard. The system is a network of flexible sensors woven or printed into a skullcap worn under a helmet. Using carbon nanotube textile nanostructures (figure 3), they have incorporated pressure sensors and flexible gyroscopes to track intensity, direction and location of impact force, as well as measure rotational motion of the head and body balance, along with lateral head motion and body balance. The cap also includes a collection of textile-based, dry sensors that measure electrical activity in the brain, including signs that may indicate the onset of mild traumatic brain injury. These sensors detect loss of consciousness, drowsiness, dizziness,

IE E Pr E oo f

384 385

C. Head and Neck Injuries

Reduction in the range of motion of the cervical spine has been found to be a useful indicator of physical disability in neck pain [70] and a predictor of poor outcome after whiplash injury [71], [72]. A new generation of wireless orientation uniaxial accelerometer and a magnetometer, in each orthogonal axis. Jasiewicz, et al. [73] evaluated IC3 orientation sensors (Intersense, Bedford, MA, USA) and showed that these wireless orientation sensors perform as well as the standard non-invasive electromagnetic devices used to measure cervical motion. These self-contained, portable and relatively inexpensive wireless sensors had very high cross-correlations

5

440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479

6

481 482 483 484 485 486 487

488

489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535

fatigue, anxiety and sensitivity to light. Finally, the skullcap includes a sensor to detect pulse rate and blood oxygen level. The system utilizes ZigBee and Bluetooth wireless telemetry to transmit data from sensors to a remote server or monitor. The goal for this type of sensor system is to provide real-time evaluation of head trauma and rapidly triage cases for conventional neuro-imaging follow-up with magnetic resonance imaging (MRI) or computerized tomography (CT). IV. C ONCLUSION AND F UTURE D IRECTIONS Recent advances in flexible electronics show great promise for healthcare monitoring. A great deal of work has been accomplished toward the integration of wearable technologies and communication [76] as well as data analysis technologies so that the goal of remote monitoring individuals in the home and community settings can be achieved. When monitoring has been performed in the home, researchers and clinicians have integrated ambient sensors in the remote monitoring systems. However some challenges remain, including efficient energy harvesting, human-device interfacing and improving the quality and range of measurements. The integration of different power sources, sensors and processing and testing in a non-controlled human environment is essential to establishing confidence in the diagnostic capabilities of these systems and their ability to change outcomes. One of the key areas of opportunity is the development and optimization of techniques to measure other physiologic metrics with higher accuracy. To date, we do not have objective, non-invasive techniques for assessing pain, mental state such as attention span and nervousness, biochemical status such as hormone levels, immunologic status for tracking exposure to infectious agents, and interactions with humans or companion animals. In addition, there is a need to develop and optimize accurate, robust and continuous physiological monitoring within closed-loop systems. For example, many non-invasive blood glucose devices have been developed, though their accuracy is limited by many confounding factors such as hydration, temperature, metabolic modulators, and comorbidities. A recent review found that 40% of continuous monitors satisfy the American Diabetes Association precision criteria of + /− 10% and fewer than 20% achieved + /− 5% accuracy. While two-thirds of these monitors were not sensitive enough to detect hypoglycemia, results from the Diabetes Control and Complications Trial suggest their potential in reducing longterm HbA1c levels [77]. A second key area of opportunity is the integration of environmental sensing and feedback to the wearer, enhancing sensory-feedback in the form of retinal and cochlear implants, and monitoring environmental risks such as radiation, pathogens, chemicals and poor air quality, and environmental parameters such as ambient temperature, lightlevels, humidity, proximity and location. With the increasing availability and reduction in cost of low power communication protocols and pervasiveness of automatic identification and data capture technologies, health care data is likely to be increasingly tagged with environmental data to describe context and enhance understanding.

A third key area of opportunity is the development of automatic algorithms that can generate relevant clinical alerts based on changes in the physiological data. Visualization tools are also needed to guide doctors, caregivers, and wearers around abnormal events, and identify increased risks that may lead to adverse health outcomes. Research toward achieving remote monitoring of older adults and subjects undergoing clinical interventions will soon face the need for establishing business models to cover the costs and identify reimbursement mechanisms for the technology and its management. Building a solid evidence base for the effectiveness of these sensor systems and addressing costs and reimbursement problems will be essential to assure that wearable sensor systems deliver on their promise of improving the quality of care for older adults and subjects affected by chronic conditions. R EFERENCES [1] J. Manyika, M. Chui, J. Bughin, R. Dobbs, P. Bisson, and A. Marrs, Disruptive Technologies: Advances that Will Transform Life, Business, and the Global Economy. McKinsey Global Institute, May 2013. [2] O. O. Ogunduyile, K. Zuva, O. A. Randle, and T. Zuva, “Ubiquitous healthcare monitoring system using integrated triaxial accelerometer, SpO2 and location sensors,” Int. J. UbiComp, vol. 4, pp. 1–13, Sep. 2013. [3] P. Bonato, “Wearable sensors/systems and their impact on biomedical engineering,” IEEE Eng. Med. Biol. Mag., vol. 22, no. 3, pp. 18–20, May/Jun. 2003. [4] P. Bonato, “Wearable sensors and systems,” IEEE Eng. Med. Biol. Mag., vol. 29, no. 3, pp. 25–36, May/Jun. 2010. [5] S. Olberding, N.-W. Gong, J. Tiab, J. A. Paradiso, and J. Steimle, “A cuttable multi-touch sensor,” in Proc. 26th Annu. ACM Symp. User Interf. Softw. Technol., 2013, pp. 245–254. [6] R. Ma et al., “Wearable 4-in. QVGA full-color-video flexible AMOLEDs for rugged applications,” J. Soc. Inf. Display, vol. 18, no. 1, pp. 50–56, 2010. [7] C. Yan et al., “Stretchable and wearable electrochromic devices,” ACS Nano, vol. 8, no. 1, pp. 316–322, Dec. 2014. [8] C. Yu, Y. Zhang, D. Cheng, X. Li, Y. Huang, and J. A. Rogers, “All-elastomeric, strain-responsive thermochromic color indicators,” Small, vol. 10, no. 7, pp. 1266–1271, 2014. [9] J. A. Rogers, T. Someya, and Y. G. Huang, “Materials and mechanics for stretchable electronics,” Science, vol. 327, no. 5973, pp. 1603–1607, Mar. 2010. [10] M. Kaltenbrunner et al., “An ultra-lightweight design for imperceptible plastic electronics,” Nature, vol. 499, pp. 458–463, Jul. 2013. [11] D.-H. Kim, J. Xiao, J. Song, Y. Huang, and J. A. Rogers, “Stretchable, curvilinear electronics based on inorganic materials,” Adv. Mater., vol. 22, no. 19, pp. 2108–2124, May 2010. [12] S. Park, M. Vosguerichian, and Z. Bao, “A review of fabrication and applications of carbon nanotube film-based flexible electronics,” Nanoscale, vol. 5, no. 5, pp. 1727–1752, 2013. [13] J. S. Park, W.-J. Maeng, H.-S. Kim, and J.-S. Park, “Review of recent developments in amorphous oxide semiconductor thin-film transistor devices,” Thin Solid Films, vol. 520, no. 6, pp. 1679–1693, Jan. 2012. [14] D.-H. Kim, R. Ghaffari, N. Lu, and J. A. Rogers, “Flexible and stretchable electronics for biointegrated devices,” Annu. Rev. Biomed. Eng., vol. 14, pp. 113–128, Aug. 2012. [15] W. Zeng, L. Shu, Q. Li, S. Chen, F. Wang, and X.-M. Tao, “Fiber-based wearable electronics: A review of materials, fabrication, devices, and applications,” Adv. Mater., vol. 26, no. 31, pp. 5310–5336, Aug. 2014. [16] A. Nathan et al., “Flexible electronics: The next ubiquitous platform,” Proc. IEEE, vol. 100, pp. 1486–1517, May 2012. [17] S. Bae et al., “Roll-to-roll production of 30-inch graphene films for transparent electrodes,” Nature Nanotechnol., vol. 5, pp. 574–578, Jun. 2010. [18] L. Hu, H. S. Kim, J.-Y. Lee, P. Peumans, and Y. Cui, “Scalable coating and properties of transparent, flexible, silver nanowire electrodes,” ACS Nano, vol. 4, no. 5, pp. 2955–2963, Apr. 2010. [19] G. Konstantatos and E. H. Sargent, “Nanostructured materials for photon detection,” Nature Nanotechnol., vol. 5, pp. 391–400, Jun. 2010.

IE E Pr E oo f

480


536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551

552 553 554 555

AQ:1

556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604

AQ:2


607 608 609 610 611 612 613 614 615

AQ:3

616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675

AQ:4

676 677 678 679

[20] M. Zhou and S. Dong, “Bioelectrochemical interface engineering: Toward the fabrication of electrochemical biosensors, biofuel cells, and self-powered logic biosensors,” Accounts Chem. Res., vol. 44, no. 11, pp. 1232–1243, Aug. 2011. [21] M. Karabulut, S. E. Crouter, and D. R. Bassett, Jr., “Comparison of two waist-mounted and two ankle-mounted electronic pedometers,” Eur. J. Appl. Physiol., vol. 95, no. 4, pp. 335–343, Oct. 2005. [22] W. Tao, T. Liu, R. Zheng, and H. Feng, “Gait analysis using wearable sensors,” Sensors, vol. 12, no. 2, pp. 2255–2283, Feb. 2012. [23] S. Chaudhuri, H. Thompson, and G. Demiris, “Fall detection devices and their use with older adults: A systematic review,” J. Geriatric Phys. Therapy, Jan. 2014. [24] B. T. Cole, S. H. Roy, C. J. De Luca, and S. H. Nawab, “Dynamical learning and tracking of tremor and dyskinesia from wearable sensors,” IEEE Trans. Neural Syst. Rehabil. Eng., vol. 22, no. 5, pp. 982–991, Sep. 2014. [25] C.-C. Yang and Y.-L. Hsu, “A review of accelerometry-based wearable motion detectors for physical activity monitoring,” Sensors, vol. 10, no. 8, pp. 7772–7788, 2010. [26] D. R. Bassett, Jr., A. V. Rowlands, and S. G. Trost, “Calibration and validation of wearable monitors,” Med. Sci. Sports Exerc., vol. 44, pp. S32–S38, Jan. 2012. [27] N. F. Butte, U. Ekelund, and K. R. Westerterp, “Assessing physical activity using wearable monitors: Measures of physical activity,” Med. Sci. Sports Exerc., vol. 44, pp. S5–S12, Jan. 2012. [28] F.-R. Fan, Z.-Q. Tian, and Z. L. Wang, “Flexible triboelectric generator,” Nano Energy, vol. 1, no. 2, pp. 328–334, Mar. 2012. [29] C. Pang et al., “A flexible and highly sensitive strain-gauge sensor using reversible interlocking of nanofibres,” Nature Mater., vol. 11, pp. 795–801, Jul. 2012. [30] T. Yamada et al., “A stretchable carbon nanotube strain sensor for human-motion detection,” Nature Nanotechnol., vol. 6, pp. 296–301, Mar. 2011. [31] D. J. Lipomi et al., “Skin-like pressure and strain sensors based on transparent elastic films of carbon nanotubes,” Nature Nanotechnol., vol. 6, pp. 788–792, Oct. 2011. [32] D. Son et al., “Multifunctional wearable devices for diagnosis and therapy of movement disorders,” Nature Nanotechnol., vol. 9, pp. 397–404, Mar. 2014. [33] S. Schreml et al., “A sprayable luminescent pH sensor and its use for wound imaging in vivo,” Experim. Dermatol., vol. 21, no. 12, pp. 951–953, Dec. 2012. [34] B. C.-K. Tee, C. Wang, R. Allen, and Z. Bao, “An electrically and mechanically self-healing composite with pressure- and flexion-sensitive properties for electronic skin applications,” Nature Nanotechnol., vol. 7, pp. 825–832, Nov. 2012. [35] D.-H. Kim et al., “Epidermal electronics,” Science, vol. 333, no. 6044, pp. 838–843, Aug. 2011. [36] J. R. Windmiller and J. Wang, “Wearable electrochemical sensors and biosensors: A review,” Electroanalysis, vol. 25, no. 1, pp. 29–46, Jan. 2013. [37] J. R. Windmiller, A. J. Bandodkar, G. Valdés-Ramírez, S. Parkhomovsky, A. G. Martinez, and J. Wang, “Electrochemical sensing based on printable temporary transfer tattoos,” Chem. Commun., vol. 48, no. 54, pp. 6794–6796, May 2012. [38] S. Bauer, “Flexible electronics: Sophisticated skin,” Nature Mater., vol. 12, pp. 871–872, Sep. 2013. [39] W.-H. Yeo et al., “Multifunctional epidermal electronics printed directly onto the skin,” Adv. Mater., vol. 25, no. 20, pp. 2773–2778, May 2013. [40] S. Xu et al., “Soft microfluidic assemblies of sensors, circuits, and radios for the skin,” Science, vol. 344, no. 6179, pp. 70–74, Apr. 2014. [41] D. Diamond, S. Coyle, S. Scarmagnani, and J. Hayes, “Wireless sensor networks and chemo-/biosensing,” Chem. Rev., vol. 108, no. 2, pp. 652–679, Jan. 2008. [42] D. Morris, S. Coyle, Y. Wu, K. T. Lau, G. Wallace, and D. Diamond, “Bio-sensing textile based patch with integrated optical detection system for sweat monitoring,” Sens. Actuat. B, Chem., vol. 139, no. 1, pp. 231–236, May 2009. [43] K. J. Fraser et al., “Wearable electrochemical sensors for monitoring performance athletes,” Proc. SPIE, Organic Semicond. Sensors Bioelectron. IV, vol. 8118, p. 81180C-1–81180C-12, Sep. 2011. [44] S. K. Vashist, “Non-invasive glucose monitoring technology in diabetes management: A review,” Anal. Chim. Acta, vol. 750, pp. 16–27, Oct. 2012.

[45] F. Michahelles, P. Matter, A. Schmidt, and B. Schiele, “Applying wearable sensors to avalanche rescue,” Comput. Graph., vol. 27, no. 6, pp. 839–847, Dec. 2003. [46] D. Curone et al., “Smart garments for safety improvement of emergency/disaster operators,” in Proc. 29th Annu. Int. Conf. IEEE Eng. Med. Biol. Soc., vol. 2007. Aug. 2007, pp. 3962–3965. [47] C. W. Mundt et al., “A multiparameter wearable physiologic monitoring system for space and terrestrial applications,” IEEE Trans. Inf. Technol. Biomed., vol. 9, no. 3, pp. 382–391, Sep. 2005. [48] A. W. Boots, J. J. B. N. van Berkel, J. W. Dallinga, A. Smolinska, E. F. Wouters, and F. J. van Schooten, “The versatile use of exhaled volatile organic compounds in human health and disease,” J. Breath Res., vol. 6, no. 2, p. 027108, Jun. 2012. [49] B. Buszewski, M. K¸esy, T. Ligor, and A. Amann, “Human exhaled air analytics: Biomarkers of diseases,” Biomed. Chromatogr., vol. 21, no. 6, pp. 553–566, Jun. 2007. [50] G. Konvalina and H. Haick, “Sensors for breath testing: From nanomaterials to comprehensive disease detection,” Accounts Chem. Res., vol. 47, no. 1, pp. 66–76, Jan. 2014. [51] L. M. L. de Lau and M. M. B. Breteler, “Epidemiology of Parkinson’s disease,” Lancet Neurol., vol. 5, no. 6, pp. 525–535, Jun. 2006. [52] N. L. W. Keijsers, M. W. I. M. Horstink, and S. C. A. M. Gielen, “Ambulatory motor assessment in Parkinson’s disease,” Movement Disorders, vol. 21, no. 1, pp. 34–44, Jan. 2006. [53] A. Salarian, H. Russmann, C. Wider, P. R. Burkhard, F. J. Vingerhoets, and K. Aminian, “Quantification of tremor and bradykinesia in Parkinson’s disease using a novel ambulatory monitoring system,” IEEE Trans. Biomed. Eng., vol. 54, no. 2, pp. 313–322, Feb. 2007. [54] S. H. Roy, B. T. Cole, L. D. Gilmore, C. J. De Luca, and S. H. Nawab, “Resolving signal complexities for ambulatory monitoring of motor function in Parkinson’s disease,” presented at the Annu. Int. Conf. IEEE Eng. Med. Biol. Soc. (EMBC), Boston, MA, USA, Aug./Sep. 2011, pp. 4832–4835. [55] V. R. Lesser, S. H. Nawab, and F. I. Klassner, “IPUS: An architecture for the integrated processing and understanding of signals,” Artif. Intell., vol. 77, no. 1, pp. 129–171, Aug. 1995. [56] Y. Balash, C. Peretz, G. Leibovich, T. Herman, J. M. Hausdorff, and N. Giladi, “Falls in outpatients with Parkinson’s disease: Frequency, impact and identifying factors,” J. Neurol., vol. 252, no. 11, pp. 1310–1315, 2005. [57] A. Weiss et al., “Can an accelerometer enhance the utility of the timed up & go test when evaluating patients with Parkinson’s disease?” Med. Eng. Phys., vol. 32, no. 2, pp. 119–125, 2010. [58] J. Cancela, M. Pastorino, M. T. Arredondo, N. S. Konstantina, F. Villagra, and M. A. Pastor, “Feasibility study of a wearable system based on a wireless body area network for gait assessment in Parkinson’s disease patients,” Sensors, vol. 14, no. 3, pp. 4618–4633, 2014. [59] J. Cancela, M. Pastorino, M. T. Arredondo, and O. Hurtado, “A telehealth system for Parkinson’s disease remote monitoring. The PERFORM approach,” in Proc. 35th Annu. Int. Conf. IEEE Eng. Med. Biol. Soc. (EMBC), Jul. 2013, pp. 7492–7495. [60] S. Patel, H. Park, P. Bonato, L. Chan, and M. Rodgers, “A review of wearable sensors and systems with application in rehabilitation,” J. Neuroeng. Rehabil., vol. 9, p. 21, Apr. 2012. [61] G. Uswatte, W. L. Foo, H. Olmstead, K. Lopez, A. Holand, and L. B. Simms, “Ambulatory monitoring of arm movement using accelerometry: An objective measure of upper-extremity rehabilitation in persons with chronic stroke,” Archives Phys. Med. Rehabil., vol. 86, no. 7, pp. 1498–1501, Jul. 2005. [62] G. Uswatte, C. Giuliani, C. Winstein, A. Zeringue, L. Hobbs, and S. L. Wolf, “Validity of accelerometry for monitoring real-world arm activity in patients with subacute stroke: Evidence from the extremity constraint-induced therapy evaluation trial,” Archives Phys. Med. Rehabil., vol. 87, no. 10, pp. 1340–1345, Oct. 2006. [63] S. K. Prajapati, W. H. Gage, D. Brooks, S. E. Black, and W. E. McIlroy, “A novel approach to ambulatory monitoring: Investigation into the quantity and control of everyday walking in patients with subacute stroke,” Neurorehabil. Neural Repair, vol. 25, no. 1, pp. 6–14, Jan. 2011. [64] S. Patel et al., “A novel approach to monitor rehabilitation outcomes in stroke survivors using wearable technology,” Proc. IEEE, vol. 98, no. 3, pp. 450–461, Mar. 2010. [65] S. L. Wolf, J. P. McJunkin, M. L. Swanson, and P. S. Weiss, “Pilot normative database for the wolf motor function test,” Archives Phys. Med. Rehabil., vol. 87, no. 3, pp. 443–445, Mar. 2006.

IE E Pr E oo f

605 606

7

680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754

8

756 757 758 759 760 761 762 763 764 765 766 767

AQ:5

768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794

[66] D. Austin, T. L. Hayes, J. Kaye, N. Mattek, and M. Pavel, “Unobtrusive monitoring of the longitudinal evolution of in-home gait velocity data with applications to elder care,” in Proc. Annu. Int. Conf. IEEE Eng. Med. Biol. Soc., vol. 2011, Aug./Sep. 2011, pp. 6495–6498. [67] S. Huang et al., “Motor impairment evaluation for upper limb in stroke patients on the basis of a microsensor,” Int. J. Rehabil. Res., vol. 35, no. 2, pp. 161–169, 2012. [68] I. Günal, N. Köse, O. Erdogan, E. Göktürk, and S. Seber, “Normal range of motion of the joints of the upper extremity in male subjects, with special reference to side,” J. Bone Joint Surgery Amer., vol. 78, no. 9, pp. 1401–1404, Sep. 1996. [69] G. Lanfermann, J. te Vrugt, A. Timmermans, E. Bongers, N. Lambert, and V. van Acht, “Philips stroke rehabilitation exerciser,” presented at the Tech. Aids Rehabil. (TAR), Aachen, Germany, Jan. 2007. [70] P. T. Dall’Alba, M. M. Sterling, J. M. Treleaven, S. L. Edwards, and G. A. Jull, “Cervical range of motion discriminates between asymptomatic persons and those with whiplash,” Spine, vol. 26, no. 19, pp. 2090–2094, 2001. [71] M. Sterling, G. Jull, B. Vicenzino, J. Kenardy, and R. Darnell, “Development of motor system dysfunction following whiplash injury,” Pain, vol. 103, nos. 1–2, pp. 65–73, 2003. [72] M. Sterling, “A proposed new classification system for whiplash associated disorders—Implications for assessment and management,” Manual Therapy, vol. 9, no. 2, pp. 60–70, 2004. [73] J. M. Jasiewicz, J. Treleaven, P. Condie, and G. Jull, “Wireless orientation sensors: Their suitability to measure head movement for neck pain assessment,” Manual Therapy, vol. 12, no. 4, pp. 380–385, 2007. [74] C. Duc, P. Salvia, A. Lubansu, V. Feipel, and K. Aminian, “A wearable inertial system to assess the cervical spine mobility: Comparison with an optoelectronic-based motion capture evaluation,” Med. Eng. Phys., vol. 36, no. 1, pp. 49–56, 2014. [75] P. Rai, S. Oh, P. Shyamkumar, M. Ramasamy, R. E. Harbaugh, and V. K. Varadan, “Nano- bio- textile sensors with mobile wireless platform for wearable health monitoring of neurological and cardiovascular disorders,” J. Electrochem. Soc., vol. 161, no. 2, pp. B3116–B3150, 2014. [76] R. Want, “iPhone: Smarter than the average phone,” IEEE Pervasive Comput., vol. 9, no. 3, pp. 6–9, Jul./Sep. 2010. [77] S. K. Vashist, “Continuous glucose monitoring systems: A review,” Diagnostics, vol. 3, no. 4, pp. 385–412, 2013.

Mary M. Rodgers received the B.S. and M.S. degrees in physical therapy from the University of North Carolina, Chapel Hill, NC, USA, in 1976 and 1981, respectively, and the Ph.D. degree in biomechanics from Pennsylvania State University, State College, PA, USA, in 1985. She is currently the George R. Hepburn Dynasplint Professor and the Vice Chair of the Department of Physical Therapy and Rehabilitation Science, University of Maryland School of Medicine, Baltimore, MD, USA, and an Advisor with the National Institute of Biomedical Imaging and Bioengineering, North Bethesda, MD, USA.

Vinay M. Pai received the B.E. degree in mechanical engineering and the M.E. degree from Saurashtra University, Rajkot, India, in 1991, the M.S.M.E. degree from the University of Iowa, Iowa City, IA, USA, in 1993, and the Ph.D. degree in mechanical engineering from Florida State University, Tallahassee, FL, USA, in 1997. He currently directs the Biomedical Imaging Informatics Program at the National Institute of Biomedical Imaging and Bioengineering, North Bethesda, MD, USA.

IE E Pr E oo f

755


Richard S. Conroy received the B.Sc. degree in laser physics and optoelectronics, and the Ph.D. degree in physics from the University of St. Andrews, St. Andrews, U.K., in 1994 and 1998, respectively, the A.L.M. degree in biotechnology from the Harvard Extension School, Cambridge, MA, USA, in 2005, and the M.I.M. degree in international management from the University of Maryland University College, Adelphi, MD, USA, in 2013. He directs the NIBIB division of applied science and technology.

795 796 797 798 799 800 801 802 803 804 805 806

807 808 809 810 811 812 813 814 815 816

817 818 819 820 821 822 823 824 825 826 827

AUTHOR QUERIES = = = = =

Please Please Please Please Please

provide the location for ref. [1]. provide the issue no. for ref. [16]. provide the volume no. and page range for ref. [23]. confirm the volume no. for refs. [43], [46], and [66]. provide the page range for ref. [69].

IE E Pr E oo f

AQ:1 AQ:2 AQ:3 AQ:4 AQ:5