A generalizable method for the assessment of static ...

A generalizable method for the assessment of static stability of walking aid users Costamagna E. 1, Thies S.B. 1, Kenney L. 1, Howard D. 2, Liu A. 1, Ogden D. 2 1

Centre for Health, Sport & Rehabilitation Sciences Research, Salford University, UK 2 School of Computing, Science and Engineering, Salford University, UK

Keywords- stability margin; pick-up walker; real-world monitoring; walking aids; biomechanics Abstract - To assist balance and mobility, older adults are often prescribed walking aids. Nevertheless, their use has been associated with increased falls-risk, although the reasons for this are, to date, entirely unknown. This problem may be addressed through an improved understanding of what exactly constitutes stable walking aid use. To this end, we present a generalizable method for the assessment of static stability of walking aid users. Our method, for the first time, considers user and device as a combined system. Our corresponding technology includes a walking frame instrumented with a load cell in each foot, force-sensing insoles, a camera system and custom-written software able to determine the stability margin of the combined system. In addition, our software informs on device dependency in terms of body weight support and timings of user and device movements. Our pilot work showed feasibility for our approach in both, laboratory and home environments. 1. INTRODUCTION Falls in older adults are a major global health problem as more than 30% of community-dwelling people aged 65 and over fall every year [1], consequences of which range from reduced activity and fear of falling, to injuries[2]. Older frail people with an unstable gait are often advised by their clinician to use walking aids, which are designed to help them maintain their balance through an increase in the effective base of support area, and through provision of structural support and haptic sensory information [3, 4]. Indeed, walking aids are used by 29-49% of older people [5]. However, paradoxically, use of walking aids (versus non-use) has been associated with a 2-fold to 3-fold risk of falling [6]. There are a number of possible explanations for this finding: one is that walking aids are prescribed to the most frail part of the population who, when falls occur, are most likely to suffer injury and hence appear in the statistics; another is that prescription of a walking aid increases the period spent upright or mobile and hence reduces time spent in a safe sitting or lying posture. Another possible explanation, and the motivation for this paper, is that incorrect device usage, either as a result of inappropriate prescription/purchasing and/or training, leads to instability and falls. Surprisingly, despite the large number of walking aid users amongst the older population, there are no generalizable methods for assessing walking aid stability. Previous work to date has either focused on the kinematics/kinetics of the user [7, 8] or their device [9-11]. Only one study collected data on user and their device (a rollator) under both, ambulatory and laboratory assessments [12]; they concluded that laboratory assessments do not fully reflect the challenges that the real world presents to walking aids users as there may be additional factors affecting their stability. Whilst their approach is praiseworthy, it remained limited in that inference on stability of the overall system (defined as person and walking aid) was not adequately addressed, leading to potentially misleading outcomes. Methods for the calculation of static stability of 6 legged robots are well established [13] and are directly applicable to this problem, yet have not been previously reported. This study aims to address this lack in the knowledge base through proposition of a novel and generalizable method for the assessment of static stability of walking aid users, based on methods from the robotics literature. Given that there are more walking frame users than users of crutches [14] and since seven times as many injuries are associated with walking frames compared with walking sticks [15], we here introduce our method for the assessment of static stability of walking frame usage. In addition, we demonstrate the application of the technology in a real world relevant setting. 2.

MATERIAL AND METHODS

I.

Combined centre of pressure relative to combined base of support as a measure of static stability

The novel method considers the user and their (four legged) walking aid as a combined system. Equation (1) shows how to calculate the position of the combined CoP: 𝑪𝑶𝑷𝒙 (𝒋) =

∑𝒏 𝒊=𝟏(𝑭𝒗𝒊 (𝒋)𝒙𝒊 ) ∑𝒏 𝒊=𝟏 𝑭𝒗𝒊 (𝒋)

,

𝑪𝑶𝑷𝒚 (𝒋) =

∑𝒏 𝒊=𝟏(𝑭𝒗𝒊 (𝒋)𝒚𝒊 ) ∑𝒏 𝒊=𝟏 𝑭𝒗𝒊 (𝒋)

(1)

where:  COPx,y are the coordinates of the CoP at time j in the mediolateral and anteroposterior direction, respectively;  Fvi is the vertical load on the ith supporting foot (either anatomical or of the frame);  xi, yi are the coordinates of the ith foot of the walking frame, or of the resultant load acting on the ith anatomical foot;  n is the number of feet in contact with the ground. When all the feet are on the ground, n=6. We define the combined CoP of user and walking frame to be the position through which the resultant load of the frame and anatomical feet acts. According to (1), at any instant in time, we therefore must know the magnitude and position of the vertical load acting on each foot of the walking frame and acting on each anatomical foot of the person. The combined BoS at each instant in time is then defined to be the line or polygon connecting each of the anatomical and/or frame feet in contact with the ground. Finally, in accordance with the walking robot literature, we define the stability margin as the shortest distance between the combined CoP and the nearest edge of the combined BoS. II.

Instrumentation development

To measure the required kinematic and kinetic data, a system was developed consisting of a) A purpose designed instrumented walking frame to measure loads acting through each of its legs b) Commercial in-shoe force sensors to measures the magnitude and location of forces acting through each anatomical foot c) An optoelectronic motion capture system (6 cameras) to capture the position of both the anatomical feet and walking frame feet. The design of a pick-up walker was modified to accommodate a single axis load cell in each leg of the frame in order to measure the vertical walker ground reaction forces (Fig1). The force data are sent to a laptop by wireless transmitters fixed onto the frame. Design requirements included the necessity to be able to adjust the frame height for a range of users, to minimise the weight added to the frame, and to assure that the axis of each load cell was perpendicular to the ground during the frame’s stance phase. Walking frame load cells, pressure insoles, and optoelectronic camera data collection were synchronised via sync pulse.

Figure 1 A) Model of an original and instrumented foot of the pick-up walker. B) Instrumented foot.

III.

Software development

In order to process the force data and estimate the stability margin of the overall system, a custom-written software algorithm was programmed in MATLAB which is able to:

    

Detect when each of the frame and user’s feet are on the ground (supporting feet) through identification of individual TD and LO events of each foot of the frame and heel strike and toe off of the user’s feet from force/insole data; the signal is lowpass filtered at 6 Hz with a 4th order Butterworth filter; Define the base of support at any time instant as the line connecting all supporting feet; Calculate the magnitude and position of the resultant load acting on each anatomical foot utilising the location of the feet and pressure data from each sensor of the insoles; Apply (1) using 3D position data and load data of the walking frame feet and the magnitude and coordinates of the resultant load that acts on each anatomical foot calculated previously ; Calculate the stability margin as the shortest distance from CoPsystem to the edges of BOSsystem.

Furthermore, the software is also able to:  Determine TD and LO sequences of the walking frame. IV.

Experimental set up

Since it is known that WFs are mostly used indoors, the WF, foot pressure, and optoelectronic motion capture systems required testing in a home-setting. For this, the systems were tested in the University of Salford Activities of Daily Living (ADL) flat, containing a bedroom, bathroom, kitchen and living room (fully furnished). 3. RESULTS All systems were proven to work in the ADL flat. Figures 2 A-B illustrate the CoP of the frame (CoPFrame), that of the user (CoPUser), and CoPSystem in relation to their respective BOS for two different time instants of the overall movement cycle. Figure 2A shows CoPFrane on the edge of BOSFrame, which, if viewed on its own, would indicate instability. However, looking at the CoPSystem it becomes clear that it is very close to CoPUser and well within the BOSSystem, therefore the overall conclusion can be that the system is stable even though the frame alone appears unstable. Similarly, in Figure 2B CoPUser is near the outer edge of BOSUser (left foot single support) due to leaning onto the device, however, as the frame provided substantial support, CoP System is well within BOSSystem and therefore one can conclude that the overall system is stable, even though the user appears unstable if viewed on their own. Finally, Figure 3 shows two examples of the timings of frame movements in relation to foot movements (correct versus incorrect according to clinical guidance). 4. DISCUSSION AND FUTURE WORK The value of this work lies in the development and demonstration of a novel method that allows for the investigation of static stability in walking aid users under real-world relevant conditions. Specifically, we have introduced a novel outcome measure, the stability margin of the combined system, which is critical for inferences on device user stability. To calculate this we record load cell, insole and 3D position data of the frame feet and anatomical feet. In addition, our system informs on the timings of user and device movements individually and in combination with one another to assess whether the movement sequence conforms with clinical guidance. Our approach is generalizable to other walking aids, including crutches and walking sticks and opens up significant opportunities to investigate stability in users of these devices.

A

B

Figure 2 Illustration of COPFrame, COPUser and COPSystem in relation to their respective BOS for four different time instants of the overall movement cycle. Figures 4A & B highlight the importance of COPSystem in relation to BOSSystem for accurate evaluation of the moving system’s stability. Black foot prints indicate stance; shaded foot prints indicate swing.

Figure 3 Movement sequence: Following clinical guidance to only move themselves once the walking frame is solid on the ground (left) versus lifting and moving the right foot whilst the walking frame is still in the air (right).

The system design presented here allows for calculation of stability margin. However, it relies on knowing the location of each CoP of each anatomical foot with respect to the frame. At present, we use position data of the anatomical and frame feet from optoelectronic cameras, but a more portable solution, based on dedicated position sensing is required for home use. In the short term, a possible way forward would be to investigate correlations between comprehensive data (including combined COP/BOS) and data we can obtain from our current system design (e.g. frame stability margin, motion of COP Frame, load transfer onto the frame and each frame foot, and LO/TD events), to infer on overall system stability from the reduced data set. 5.

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11.

12.

13. 14. 15.

REFERENCES

Rubenstein, L.Z., Falls in older people: epidemiology, risk factors and strategies for prevention. Age Ageing, 2006. 35 Suppl 2: p. ii37-ii41. Bloem, B.R., J.A. Steijns, and B.C. Smits-Engelsman, An update on falls. Curr Opin Neurol, 2003. 16(1): p. 15-26. Jeka, J.J., Light Touch Contact as a Balance Aid. Physical Therapy, 1997. 77(5): p. 476-487. Maeda, A., et al., Postural sway during cane use by patients with stroke. Am J Phys Med Rehabil, 2001. 80(12): p. 903-8. Lofqvist, C., et al., Assistive devices among very old people in five European countries. Scand J Occup Ther, 2005. 12(4): p. 181-92. Deandrea, S., et al., Risk factors for falls in community-dwelling older people: a systematic review and meta-analysis. Epidemiology, 2010. 21(5): p. 658-68. Kloos, A.D., et al., The impact of different types of assistive devices on gait measures and safety in Huntington's disease. PLoS One, 2012. 7(2): p. e30903. Kuan, T.S., J.Y. Tsou, and F.C. Su, Hemiplegic gait of stroke patients: the effect of using a cane. Arch Phys Med Rehabil, 1999. 80(7): p. 777-84. Pardo, R.D., A.B. Deathe, and D.A. Winter, Walker user risk index. A method for quantifying stability in walker users. Am J Phys Med Rehabil, 1993. 72(5): p. 301-5. Sardini, E., et al., Wireless Instrumented Crutches for Force and Tilt Monitoring in Lower Limb Rehabilitation. Procedia Engineering, 2014. 87: p. 348-351. W. Wu, L.A., B. Jordan, T. Stathopoulos, M. Batalin, and A.V. W. Kaiser, M. Sarrafzadeh, M. Fang, J. Chodosh, The SmartCane System An Assistive Device for Geriatrics. geriatrics. Third International Conference on Body Area Net-works (BodyNets 2008), 2008. Tung, J.Y., et al., Combining ambulatory and laboratory assessment of rollator use for balance and mobility in neurologic rehabilitation in-patients. Disabil Rehabil Assist Technol, 2015. 10(5): p. 40714. Liu, A., The mechanical design of legged robots , PhD thesis, University of Salford, in School of health care professions. 2009, University of Salford. Kaye, H.S., T. Kang, and M.P. LaPlante, Mobility device use in the United States. Vol. 14. 2000. Stevens, J.A., et al., Unintentional fall injuries associated with walkers and canes in older adults treated in U.S. emergency departments. J Am Geriatr Soc, 2009. 57(8): p. 1464-9.

6. ACKNOWLEDGEMENTS The authors wish to acknowledge funding for this work from the University of Salford’s Pathways to Excellence scheme.