Magn Reson Med Sci
© 2015 Japanese Society for Magnetic Resonance in Medicine
MAJOR PAPER
E-pub ahead of print by J-STAGE doi:10.2463/mrms.2014-0086
Optimization of Inversion Time for Postmortem Fluid-attenuated Inversion Recovery (FLAIR) MR Imaging at 1.5T: Temperature-based Suppression of Cerebrospinal Fluid Kazuyuki ABE1*, Tomoya KOBAYASHI2, Seiji SHIOTANI3, Hajime SAITO2, Kazunori KAGA2, Kazuya TASHIRO2, Satoka SOMEYA2, Hideyuki HAYAKAWA4, and Kazuhiro HOMMA5 1
Department of Radiological Science, Faculty of Health Sciences, Junshin Gakuen University 1–1–1 Chikushigaoka, Minami-ku, Fukuoka 815–8510, Japan 2 Department of Radiological Technology, Tsukuba Medical Center Hospital 3 Department of Radiology, Tsukuba Medical Center Hospital 4 Department of Forensic Medicine, Tsukuba Medical Examiner’s Office 5 The National Institute of Advanced Industrial Science and Technology, Human Technology Research Institute (Received August 18, 2014; Accepted December 10, 2014; published online March 31, 2015)
Purpose: Signal intensity (SI) and image contrast on postmortem magnetic resonance (MR) imaging are different from those of imaging of living bodies. We sought to suppress the SI of cerebrospinal fluid (CSF) sufficiently for fluid-attenuated inversion recovery (FLAIR) sequence in postmortem MR (PMMR) imaging by optimizing inversion time (TI). Materials and Methods: We subject 28 deceased patients to PMMR imaging 3 to 113 hours after confirmation of death (mean, 27.4 hrs.). PMMR imaging was performed at 1.5 tesla, and T 1 values of CSF were measured with maps of relaxation time. Rectal temperatures (RT) measured immediately after PMMR imaging ranged from 6 to 32°C (mean, 15.4°C). We analyzed the relationship between T 1 and RT statistically using Pearson’s correlation coefficient. We obtained FLAIR images from one cadaver using both a TI routinely used for living bodies and an optimized TI calculated from the RT. Results: T 1 values of CSF ranged from 2159 to 4063 ms (mean 2962.4), and there was a significantly positive correlation between T 1 and RT (r = 0.96, P < 0.0001). The regression expression for the relationship was T 1 = 74.4 * RT + 1813 for a magnetic field strength of 1.5T. The SI of CSF was effectively suppressed with the optimized TI (0.693 * T 1 ), namely, TI = 0.693 * (77.4 * RT + 1813). Conclusion: Use of the TI calculated from the linear regression of the T1 and RT optimizes the FLAIR sequence of PMMR imaging. Keywords: body temperature, fluid-attenuated inversion recovery (FLAIR), postmortem cross-sectional imaging, postmortem magnetic resonance (PMMR) imaging, T1 value
Introduction The worldwide decline in the rate of conventional autopsies 1,2 has increased the need for and frequency of postmortem imaging as a complementary, supplementary, or alternative method of autopsy *Corresponding author, Phone: +81-92-554-1255, Fax: +8192-552-2707, E-mail:
[email protected]
method. 3–17 Interpreting the relationship of imaging findings to cause of death requires knowledge of normal postmortem imaging findings. However, consensus regarding what constitutes a normal appearance on PMMR images is difficult because T 1 and T 2 values change according to body temperature. 12,18 –20 Therefore, acquisition of appropriate image contrast for deceased bodies at low temperatures re-
K. Abe et al.
quires optimization of the parameters for PMMR imaging. Short-tau inversion recovery (STIR) and fluidattenuated inversion recovery (FLAIR) are typical sequences employed to suppress tissue signal. 21 Application of the inversion time (TI) for living bodies for PMMR imaging does not adequately suppress fat signals in STIR and CSF signals in FLAIR.18 –20 Suppression of the signal intensity (SI) of fat has been reported with STIR in PMMR imaging when using an optimized TI based on the measured rectal temperature (RT). 22 Therefore, we surmised that the SI of CSF can also be suppressed on postmortem FLAIR images by adjusting the TI according to the T 1 temperature dependence of water. 18,20 In this study, we attempted to achieve sufficient SI suppression of the CSF on cerebral FLAIR PMMR images by optimizing the TI.
Methods Subjects Our subjects were 28 deceased patients (18 male,10 female) aged 9 months to 92 years (mean, 59.2 years) who underwent PMMR imaging and autopsy. In all subjects, the autopsy performed at the medical examiner’s office after PMMR imaging proved no abnormalities in the brain, such as trauma, cerebrovascular disease, tumor, or inflammation. Causes of death included 7 cases of ischemic heart disease, five of sudden cardiac arrest, 2 cases each of drug toxicity, malnutrition, alcoholic liver disease, and hypothermia, and one case each of gastric perforation, rupture of right cardiac chambers, prostate abscess, ovarian tumor, burn shock, tension pneumothorax, suffocation, and aortic injury.
utilizes the variable flip angle for a 3-dimensional volumetric interpolated breath hold examination (3D-VIBE). 24 Scan parameters were: repetition time (TR), 25 ms; echo time (TE), 1.79 ms; flip angle, 3° and 15°; pixel size, 0.9 mm © 0.9 mm; slice thickness, 3 mm; number of slices, 22; and scan time, 269 s. We obtained FLAIR images of a 66-year-old woman who died of aortic injury, 19 hours after death using the TI routinely used for living bodies (2300 ms) and a TI optimized for cadavers with low temperature. The RT of the subject was 16°C. We calculated the T 1 value using the derived regression of the other 27 subjects, excluding the 66year-old female subject, and calculated the optimized TI by multiplying the T 1 by 0.693 times (TI = 0.693 * T 1 value). 21 Other scan parameters were: TR, 9000 ms; TE, 106 ms; echo train length, 21; pixel size, 1.0 mm © 0.9 mm; slice thickness, 6 mm; gap, = one mm; number of slices, 20; and scan time, 180 s. A board-certified radiologist compared the CSF signals of the 2 FLAIR images. Statistical analysis We placed circular regions of interest (ROI) of 3 mm diameter in areas of CSF in the anterior horns of bilateral ventricles of the brain of each subject and measured the T 1 value in each ROI (Fig. 1). We analyzed the relationship between the average T 1 and RT using Pearson’s correlation coefficient, obtaining the coefficient using the least squares method, and derived the linear regression between RT and T 1 and used it to calculate the optimal TI nec-
Examination setting The subjects were kept in cold storage at 4°C for 3 to 113 hours after death (mean 27.4 hrs.) and then subjected to PMMR imaging by permission of the ethics committee of our institution. RTs measured immediately after PMMR imaging using an industrial thermometer (7-257-01, As One Corporation, Osaka, Japan) ranged from 6 to 32°C (mean, 15.4°C). MR imaging PMMR imaging was performed using a 1.5 tesla clinical scanner with an 8-channel head coil (Magnetom Avanto, Siemens, Erlangen, Germany). T1 values were measured using a relaxation time map creation tool (syngo MapIt, Siemens), 23 which E-pub ahead of print
Fig. 1. Regions of interest (ROI) used to measure T1 values from T1 maps. Two circular ROIs are set in areas of cerebrospinal fluid (CSF) in the anterior horns of bilateral ventricles of the brain (arrows). Magnetic Resonance in Medical Sciences
Optimization of Postmortem FLAIR MR Imaging
essary to nullify the CSF signal by multiplying the T 1 by 0.693. 21 Statistical analysis was performed using Excel 2010 (Microsoft, Redmond, WA, USA) with Statcel 2 add-in software (OMS, Tokyo, Japan).
Results T 1 values of the CSF in the lateral ventricle did not vary between the left and right sides (P = 0.99), so we averaged the values from the 2 sides. T 1 values of the lateral ventricle ranged from 2159 to 4063 ms (mean « standard deviation, 2962.4 « 624.4 ms). Figure 2 shows the relationship between the T 1 values of the CSF in the lateral ventricle and RT. There was a significant correlation between T 1
and RT (r = 0.96, P < 0.0001). Based on the data of the 27 subjects excluding that of the 66-year-old woman who died of aortic injury for whom we employed 2 different parameters, we determined the derived regression expression as: T 1 = 74.5 * RT + 1818. Utilizing this equation and the RT of 16°C of the 66-year-old woman, we first calculated the estimated T 1 as 74.5 * 16 + 1818 = 3010 ms and then the optimized TI as 0.693 * the estimated T 1 = 0.693 * 3010 Ü 2086 ms. The CSF signal was effectively suppressed using the optimized TI for cadavers compared to the routine TI for living bodies (Fig. 3). After reconsiderin inclusionof the results of the 66-year-old subject, we conceived the following regression expressions for all 28 subjects: T1 = 74.4 * RT + 1813, and TI = 0.693 * T 1 . Hence, TI = 0.693 * (74.4 * RT + 1813).
Discussion
Fig. 2. Correlation between T1 values of cerebrospinal fluid (CSF) and rectal temperature. The solid line represents the result of linear regression.
In clinical practice, FLAIR imaging is useful to suppress CSF signals, allowing differentiation of lesions in the cerebral parenchyma, such as those of encephalitis, cerebral contusion, cerebral infarction, and subarachnoid hemorrhage. 25 –28 However, when postmortem examination employs the TI routinely used for living bodies, lesion detection by FLAIR imaging is diminished by unclear image contrast 18,19 caused by insufficient signal suppression of the CSF. Thus, parameters for FLAIR of
Fig. 3. Axial fluid-attenuated inversion recovery (FLAIR) images of a case of aortic injury (rectal temperature, 16°C). (a) FLAIR obtained with a routine inversion time (TI) for living bodies (2300 ms). (b) FLAIR obtained with an optimized TI for cadavers (TI = 0.693 * T1 = 0.693 * (74.5 * 16 + 1818) Ü 2086 ms). The signal of cerebrospinal fluid (CSF) was significantly suppressed using the TI optimized for cadavers with cold temperature compared to a routine TI for living bodies. Magnetic Resonance in Medical Sciences
E-pub ahead of print
K. Abe et al.
PMMR imaging should be optimized for deceased subjects. Tofts and associates reported that in cadaveric brains, T 1 values of the CSF can be derived by measuring the diffusion coefficient of the CSF at low temperatures, and CSF signals can then be suppressed effectively by using an optimized TI. 18 However, such a method is applicable only when the CSF is in a normal oxyhemoglobin changes the ADC values of the CSF. 29 Optimization of STIR PMMR imaging using the TI, which is calculated based on RT, allows sufficient suppression of the SI of fat, 22 and we observed effective suppression of the CF signal in FLAIR images by adapting this method. Inadequate suppression of the cadaveric CSF signal even following optimization using this method may indicate the CSF is abnormal, such as in the case of CSF containing blood, findings similar to the abnormal findings in a living body. 30,31 Further study is necessary to evaluate the rate of lesion detection using optimized parameters for FLAIR in PMMR imaging. When using a physiologically normal body temperature of 37°C, T 1 values of the CSF at 1.5T differed greatly with the application of our equation for cadavers (T 1 = 74.4 * RT + 1813). 21 The theoretically calculated T 1 value of CSF is 4500 ms, although the actual T 1 of a living body is approximately 3600 ms. This discrepancy would be explained by pulsation and oxygen concentration in CSF. The T 1 value is shorter in accordance with the faster pulsation 32 and higher oxygen concentration in CSF. 33–35 The null pulsation of CSF and reduced levels of oxygen concentration of cadavers result in prolonged T 1 values compared with those of living bodies. One limitation of our study was that the equation, T 1 = 74.4 * RT +813, is effective only for a magnetic field strength of 1.5T. Though such an equation for different magnetic field strengths would also be expressed as a linear equation, relaxation times differ at different magnetic field strengths. In conclusion, we observed significant correlation between the T 1 of CSF and the RT in PMMR, which is expressed by the linear equation T 1 = 74.4 * RT + 1813. TI optimization for PMMR FLAIR imaging, namely, TI = 0.693 * (74.4 * RT + 1813), is considered feasible to suppress CSF signals.
Acknowledgement This work was supported by a grant from the Daiwa Securities Health Foundation. E-pub ahead of print
References 1.
2. 3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
Burton JL, Underwood J. Clinical, educational, and epidemiological value of autopsy. Lancet 2007; 369: 1471–1480. Shojania KG, Burton EC. The vanishing nonforensic autopsy. N Engl J Med 2008; 358:873–875. Brogdon BG. Research and applications of the new modalities. In: Brogdon BG, ed. Forensic Radiology. 1st ed. Boca Raton: CBC, 1998; 333–338. Swift B, Rutty GN. Recent advances in postmortem forensic radiology–computed tomography and magnetic resonance imaging applications. In: Tsokos M, ed. Forensic Pathology Reviews. 1st ed. Totowa: Humana, 2006; 355–404. Oesterhelweg L, Thali MJ. Experiences with virtual autopsy approach worldwide. In: Thali MJ, Dirnhofer R, Vock P, eds. The virtopsy approach. 1st ed. Boca Raton: CRC, 2009; 475–477. Kaneko T, Hibi M, Ishibashi M, et al. Postmortem computed tomography is an informative approach for prevention of sudden unexpected natural death in the elderly. Risk Manag Healthc Policy 2010; 3: 13–20. Roberts IS, Benamore RE, Benbow EW, et al. Postmortem imaging as an alternative to autopsy in the diagnosis of adult deaths: a validation study. Lancet 2012; 379:136–142. Wichmann D, Obbelode F, Vogel H, et al. Virtual autopsy as an alternative to traditional medical autopsy in the intensive care unit: a prospective cohort study. Ann Intern Med 2012; 156:123–130. Takahashi N, Higuchi T, Shiotani M, et al. The effectiveness of postmortem multidetector computed tomography in the detection of fatal findings related to cause of non-traumatic death in the emergency department. Eur Radiol 2012; 22:152–160. Rutty GN, Brogdon G, Dedouit F, et al. Terminology used in publications for post-mortem cross-sectional imaging. Int J Legal Med 2013; 127:465– 466. Okuda T, Shiotani S, Sakamoto N, Kobayashi T. Background and current status of postmortem imaging in Japan: short history of “autopsy imaging (Ai)”. Forensic Sci Int 2013; 225:3–8. Ruder TD, Thali MJ, Hatch GM. Essentials of forensic post-mortem MR imaging in adults. Br J Radiol 2014; 87:20130567. Flach PM, Thali MJ, Germerott T. Times have changed! Forensic radiology–a new challenge for radiology and forensic pathology. AJR Am J Roentgenol 2014; 202:W325–334. Morgan B, Adlam D, Robinson C, Pakkal M, Rutty GN. Adult post-mortem imaging in traumatic and cardiorespiratory death and its relation to clinical radiological imaging. Br J Radiol 2014; 87:20130662. De Groot CJ, Bergers E, Kamphorst W, et al. Postmortem MRI-guided sampling of multiple sclerosis brain lesions: increased yield of active demyelinating and (p)reactive lesions. Brain 2001; 124:1635–1645. Magnetic Resonance in Medical Sciences
Optimization of Postmortem FLAIR MR Imaging
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
Bö L, Geurts JJ, Ravid R, Barkhof F. Magnetic resonance imaging as a tool to examine the neuropathology of multiple sclerosis. Neuropathol Appl Neurobiol 2004; 30:106–117. Kanawaku Y, Someya S, Kobayashi T, et al. Highresolution 3D-MRI of postmortem brain specimens fixed by formalin and gadoteridol. Leg Med (Tokyo) 2014; 16:218–221. Tofts PS, Jackson JS, Tozer DJ, et al. Imaging cadavers: cold FLAIR and noninvasive brain thermometry using CSF diffusion. Magn Reson Med 2008; 59: 190–195. Kobayashi T, Shiotani S, Kaga K, et al. Characteristic signal intensity changes on postmortem magnetic resonance imaging of the brain. Jpn J Radiol 2010; 28:8–14. Kobayashi T, Isobe T, Shiotani S, et al. Postmortem magnetic resonance imaging dealing with low temperature objects. Magn Reson Med Sci 2010; 9:101– 108. Hashemi RH, Bradley WG, Lisanti CJ. Tissue suppression techniques. In: Hashemi RH, Bradley WG, Lisanti CJ, eds. MRI: the basics. 3rd ed. Philadelphia, Pa: Lippincott Williams & Wilkins, 2010; 284– 288. Kobayashi T, Monma M, Baba T, et al. Optimization of inversion time for postmortem short-tau inversion recovery (STIR) MR imaging. Magn Reson Med Sci 2014; 13:67–72. Kobayashi T, Ookubo J, Monma M, et al. [Evaluation of measurement accuracy of a T1 and T2 mapping tool]. Jpn J Magn Reson Med 2012; 32:66–75. [Article in Japanese with English abstract] Wang HZ, Riederer Step J, Lee JN. Optimizing the precision in T1 relaxation estimation using limited flip angles. Magn Reson Med 1987; 5:399– 416. White ML, Edwards-Brown MK. Fluid attenuated inversion recovery (FLAIR) MRI of herpes encephalitis. J Comput Assist Tomogr 1995; 19:501–502. Ashikaga R, Araki Y, Ishida O. MRI of head injury
Magnetic Resonance in Medical Sciences
27.
28.
29.
30.
31. 32.
33.
34.
35.
using FLAIR. Neuroradiology 1997; 39:239–242. Noguchi K, Ogawa T, Inugami A, et al. MRI of acute cerebral infarction: a comparison of FLAIR and T2weighted fast spin-echo imaging. Neuroradiology 1997; 39:406– 410. Tsuchiya K, Inaoka S, Mizutani Y, Hachiya J. Fast fluid-attenuated inversion-recovery MR of intracranial infections. AJNR Am J Neuroradiol 1997; 18: 909–913. Atlas SW, DuBois P, Singer MB, Lu D. Diffusion measurements in intracranial hematomas: implications for MR imaging of acute stroke. AJNR Am J Neuroradiol 2000; 21:1190–1194. Stuckey SL, Goh TD, Heffernan T, Rowan D. Hyperintensity in the subarachnoid space on FLAIR MRI. AJNR Am J Neuroradiol 2007; 189:913–921. Kidwell CS, Wintermark M. Imaging of intracranial haemorrhage. Lancet Neurol 2008; 7:256–267. Bakshi R, Caruthers SD, Janardhan V, Wasay M. Intraventricular CSF pulsation artifact on fast fluid-attenuated inversion-recovery MR images: analysis of 100 consecutive normal studies. AJNR Am J Neuroradiol 2000; 21:503–508. Deliganis AV, Fisher DJ, Lam AM, Maravilla KR. Cerebrospinal fluid signal intensity increase on FLAIR MR images in patients under general anesthesia: the role of supplemental O2. Radiology 2001; 218:152–156. Braga FT, da Rocha AJ, Hernandez Filho G, Arikawa RK, Ribeiro IM, Fonseca RB. Relationship between the concentration of supplemental oxygen and signal intensity of CSF depicted by fluid-attenuated inversion recovery imaging. AJNR Am J Neuroradiol 2003; 24:1863–1868. Frigon C, Shaw DW, Heckbert SR, Weinberger E, Jardine DS. Supplemental oxygen causes increased signal intensity in subarachnoid cerebrospinal fluid on brain FLAIR MR images obtained in children during general anesthesia. Radiology 2004; 233: 51–55.
E-pub ahead of print
