Forensic Sci Med Pathol (2013) 9:363–366 DOI 10.1007/s12024-013-9456-0
ORIGINAL ARTICLE
Magnetic resonance imaging (MRI) of bruises: a pilot study Neil E. I. Langlois • Claire G. Ross Roger W. Byard
•
Accepted: 3 May 2013 / Published online: 13 June 2013 Ó Springer Science+Business Media New York 2013
Abstract The purpose of this study was to investigate if magnetic resonance imaging (MRI) could be used to image the presence of hemosiderin in bruises and if there was the potential for this technique to be applied as a non-invasive method to estimate the age of bruises. To achieve this aim an animal model to produce lesions resembling bruises was created by injecting blood obtained from the tail vein subcutaneously into an area of the abdominal wall. The animals were euthanized at 3, 6, 12 h, 1, 2, 3, 5, and 7 days post injection and the skin of the abdominal wall was excised for MRI scanning and histological examination. The injected blood appeared as hypointense (dark) areas on the T2* MRI at 3 and 6 h. The image of the injected areas became indistinct at 12 h and continued to be indistinct at 1 and 2 days, although there appeared to be transitioning from hypointensity to hyperintensity (light). The magnetic resonance image appeared to better correspond to the histological appearance at 3 and 5 days, with the ‘‘bruise’’ appearing hyperintense (white); however, some hypointense (darker) areas at 3 day possibly corresponded to the development of hemosiderin. At 7 day the injected blood had been converted to hemosiderin with possible correlation between areas of blue staining in Perls’ stained histologic sections and areas of extreme hypointensity in the T2* magnetic resonance image. This study has shown that a series of changes occur on MRI of bruises in an animal model that may relate to histological changes. Although N. E. I. Langlois C. G. Ross R. W. Byard The University of Adelaide School of Medical Sciences, Frome Road, Adelaide, SA 5005, Australia N. E. I. Langlois (&) C. G. Ross R. W. Byard Forensic Science SA, 21 Divett Place, Adelaide, SA 5000, Australia e-mail:
[email protected]
variability in the intensity of the MRI signal and considerable soft tissue artifact currently make interpretations difficult, this may be a technique worth pursuing in the non-invasive evaluation of bruises. Keywords Bruise Hemosiderin Magnetic resonance imaging Animal model Time
Introduction Estimation of the age of bruises is a contentious area [1, 2]. In forensic autopsy work histological assessment of hemosiderin may be used as an adjunctive method to indicate that a bruise is not recent [3]. Hemosiderin is formed from iron that is released when blood is broken down [4, 5] and is readily displayed in histological sections using Perls’ Prussian blue stain [6, 7]. However, it has not been established with accuracy when stainable hemosiderin first appears in bruises in humans, or how long it persists in human subjects [8]. This cannot be investigated by taking serial biopsy samples from living volunteers, as obtaining the biopsy creates a wound that will result in hemosiderin deposition; thus, a non-invasive method is required. Visual inspection of bruises is not possible for this purpose, as yellow coloration from hemosiderin [5] may be obscured by yellow coloration from bilirubin [9, 10] that is also produced from the breakdown of red blood cells at the bruise site [1]. In addition, visual inspection of bruises will not provide information on deeper tissue levels and may not reveal changes at the true site of trauma, as opposed to the area where blood is merely visible at the skin surface [11, 12]. Spectrophotometry is unlikely to assist with this issue, as hemosiderin has a non-specific absorption spectrum that is similar to melanin [13]. As it has been reported
123
364
that magnetic resonance imaging (MRI) can detect the presence of hemosiderin in intracranial scans [14–16] this study was undertaken to investigate the potential application that MRI may have for the detection of hemosiderin in bruises in an animal model.
Methods An animal model to produce lesions resembling bruises was created by subcutaneously injecting blood obtained from the tail vein into a region of the abdominal wall of eight adult male Sprague–Dawley rats. 0.1 ml of blood was withdrawn and re-injected subcutaneously under general anesthesia (Isoflurane 0.5–5 % delivered through an oxygen flow system). The rats were then allowed to recover and were euthanized by carbon dioxide gas and cervical dislocation at 3, 6, 12 h, 1, 2, 3, 5, and 7 days post injection. The skin of the abdominal wall was excised for MRI scanning and histological examination by staining with hematoxylin and eosin (H&E) as well as Perls’ Prussian blue for hemosiderin. The study was approved by the University of Adelaide Animal Ethics Committee (M-2012-029), the SA Pathology/CHN Animal Ethics Committee (02/12) and the Forensic Science R&D committee. Imaging was performed at the National Imaging Facility, Large Animal Research and Imaging Facility (LARIF), Gilles Plains, using a 1.5 T Siemens Sonata MR scanner and Platform Technology Development Officer Software (NUMARIS/4 version syngo MR 2004A). The excised skin was embedded flat within 40 % gelatin, in a round 10 cm diameter, 1.5 cm deep clear plastic container. This was then placed directly underneath a Human Temporal Mandibular Joint (TMJ) LP loop coil (one element of the Siemens double loop array). A series of T2* axial cuts were obtained using the following parameters: Gradient Echo (GE)-T2star-map sequence with repetition time (TR) = 300 ms, echo time (TE) = 7.5 ms, flip angle 20°, field-of-view (FoV) = 150 9 150 mm, image matrix = 256 9 256, slice thickness 2.5 mm (Axial), slice gap = 0.45 mm. Images were exported as Dicom files and viewed using OsiriXÒ (v4.1.2 64-bit). After the MRI scanning had been completed the tissue was fixed in 10 % buffered formalin. Histological sections were examined following staining with hematoxylin and eosin (H&E) and Perls’ Prussian blue (2 % aqueous potassium ferrocyanide solution and 2 % aqueous hydrochloric acid solution counterstained using stock safranin solution 1 %). Comparison between the MR images and the histology was performed visually by all authors (NEIL and then RWB with CGR independently).
123
Forensic Sci Med Pathol (2013) 9:363–366
Results The injected blood appeared to be represented by extreme hypointensity (black areas) on the T2* MRI at 3 and 6 h. The image of the injected area became indistinct at 12 h; it continued to be indistinct at 1 and 2 days, but appeared to be transitioning from hypointensity (dark) to hyperintensity (light). The magnetic resonance image appeared to better correspond to the histological appearance at 3 and 5 days, with the ‘‘bruise’’ appearing hyperintense; however, some hypointense areas possibly appeared at 3 days, corresponding to the development of definite staining for hemosiderin. At 7 days the injected blood had been converted to hemosiderin with apparent correlation between the blue staining in the Perls’ sections and extreme hypointensity in the T2* magnetic resonance image (see Fig. 1).
Discussion The presence of hemorrhage within bruised tissues has been demonstrated previously by MRI in a rat model using an impact method to produce the bruise [17]. For this study an animal model to create focal collections of subcutaneous blood with consequent inflammation resulting in red cell breakdown and formation of hemosiderin was developed; injection of blood rather than blunt impact was used to minimize tissue trauma to the animal. Histological assessment of the injection sites showed that blood was present in a focal hematoma at 3 h, but had clearly diffused into adjacent tissues by 12 h. Stainable hemosiderin was focally present at 1 day, but was much more clearly defined by 3 days, which is in keeping with the observation that hemosiderin deposits may be seen after 1 day in the mouse model [4]. By 7 days there was no blood remaining around the injection site, but hemosiderin was clearly evident on the Perls’ stained sections. Thus, although the lesions used in the animal model were not true bruises they appeared to undergo the changes that would be expected in a bruise. Hemosiderin, which contains oxidized iron (Fe3?), has paramagnetic properties compared to blood (which contains Fe2?) resulting in extreme hypointensity (black areas) on T2* images [18–20]. This appears to have been imaged in the 7 day sample. However, any substance that causes local variation in the strength of the magnetic field can produce signal changes that can appear as areas of hypointensity. Deoxyhemoglobin can exert paramagnetic effects due to the presence of unpaired electrons. This could account for the appearance of the images of the early time points (3 and 6 h). The cause of the apparent transition from hypointensity to hyperintensity in the samples from 12 h to 5 days remains to be elucidated. Methemoglobin and ferritin also exert paramagnetic effects due to
Forensic Sci Med Pathol (2013) 9:363–366
365
the presence of Fe3? [14, 15]. Furthermore, a chemical shift effect can occur between fat and water interfaces [21, 22], resulting in dark areas that can appear similar to the effect caused by hemosiderin. Factors such as these may account for the imperfect correlation between MR images and histology. Other influences, such as shrinkage and distortion of tissue during fixation may also have impaired the correlation. Air can also cause rapid dephasing of the T2 signal in the same manner as hemosiderin [22–24]. This can appear as hypo intensity at the air/skin interface, which can be seen in Fig. 1. Thus, the surface air–skin interface poses a challenge for the application of MRI to bruises in the clinical situation in human subjects. Nonetheless, with the introduction of high field strength (7 T) clinical scanners and high resolution surface imaging coils MRI may become a useful tool for estimating the age of bruises and examining the persistence of hemosiderin, given the results of this pilot study. In summary, this study has shown that alterations in the histologic features of bruises over time may be accompanied by alterations in MRI. Correlating these two methods of assessment will be challenging.
Key points 1. 2.
3.
4.
Injecting blood subcutaneously in a rat model can simulate a bruise- with local formation of hemosiderin. It appeared possible to obtain images of the injected blood using a T2* sequence on a 1.5T MRI scanner as extreme hypointensity (black regions) at 3 and 6 h after injection. There was also hypointensity on the T2* magnetic resonance image that may have been due to the presence of hemosiderin. A comparison of magnetic resonance images and histology suggests that MRI may have a role in noninvasively estimating the age of bruises.
Acknowledgments The authors acknowledge the work and support of Ms Maria Bellis, Dr Gregory Brown, Dr Gary Cowin, Ms Diana Pilkington, Ms Jayne Skinner, Ms Tasma How, and Ms Melissa Walker.
References
Fig. 1 Composite image for comparison of T2* 1.5T MRI axial images (left) with histology (right) at time points 3 h to 7 days (3 h to 3 days—hematoxylin and eosin, H&E; 5 and 7 day—Perl’s stain)
1. Langlois NEI. The science behind the quest to determine the age of bruises—a review of the English language literature. Forensic Sci Med Pathol. 2007;3(4):241–51. 2. Byard RW, Wick R, Gilbert JD, Donald T. Histologic dating of bruises in moribund infants and young children. Forensic Sci Med Pathol. 2008;4:187–92.
123
366 3. Vanezis P. Intrepreting bruises at necropsy. J Clin Pathol. 2001;54:348–55. 4. Muir R, Niven JSF. The local formation of blood pigments. J Pathol. 1935;41:183–97. 5. Ritcher GW. The iron-loaded cell—the cytopathology of iron storage. A review. Am J Pathol. 1978;91:362–404. 6. Wixom RL, Prutkin L, Munro HN. Hemosiderin: nature, formation, and significance. Int J Exp Path. 1980;22:193–225. 7. Bancroft JD, Cook HC. Manual of histological techniques and their diagnostic application. 1st ed. Edinburgh: Churchill Livingstone; 1984. 8. Hughes VK, Langlois NEI. Visual and spectrophotometric observations related to histology in a small sample of bruises from cadavers. Forensic Sci Med Pathol. 2011;7:253–6. 9. Lemberg R, Wyndham RA. Reduction of biliverdin to bilirubin in tissues. Biochem J. 1936;30:1147–70. 10. Houlihan DD, Armstrong MJ, Newsome PN. Investigation of jaundice. Medicine. 2011;39:518–22. 11. Kaczor K, Pierce MC, Makoroff K, Corey TS. Bruising and physical child abuse. Clin Pediatr Emerg Med. 2006;7:153–60. 12. Nash KR, Sheridan DJ. Can one accurately date a bruise? State of the science. J Forensic Nurs. 2009;5:31–7. 13. Wells CL, Wolken JJ. Microspectrophotometry of haemosiderin granules. Nature. 1962;193:977–8. 14. Bradley WG. MR appearance of hemorrhage in the brain. Radiology. 1993;189:15–26.
123
Forensic Sci Med Pathol (2013) 9:363–366 15. Gomori JM, Grossman RI. Mechanisms responsible for the MR appearance and evolution of intracranial hemorrhage. Radiographics. 1988;8:427–40. 16. Messori A, Salvolini U. Postmortem MRI as a useful tool for investigation of cerebral microbleeds. Stroke. 2003;34:376–7. 17. Sherman D, Bir C, Viano D, Haacke EM, editors. Evaluation and quantification of bruising. ISB XXth Congress—ASB 29th annual meeting. Cleveland, Ohio; 2005. 18. Imaizumi T, Chiba M, Honma T, Niwa J. Detection of hemosiderin deposition by T2*-weighted MRI after subarachnoid hemorrhage. Stroke. 2003;34:1693–8. 19. Narva´ez JA, Narva´ez J, Ortega R, De Lama E, Roca Y, Vidal N. Hypointense synovial lesions on T2-weighted images: differential diagnosis with pathologic correlation. Am J Radiol. 2003;181:761–9. 20. Kim J, Dembner AG. Renal hemosiderosis. Appl Radiol. 2008;37:56D-E. 21. Scherzinger AL, Hendee WR. Basic principles of magnetic resonance imaging. An update. West J Med. 1985;143:782–92. 22. Foltz WD, Jaffray DA. Principles of magnetic resonance imaging. Radiat Red. 2012;177:331–48. 23. Storey P. Introduction to magnetic resonance imaging and spectroscopy. Methods Mol Med. 2006;124:3–57. 24. Chanvman GB, Babyn PS, Thomas B, Shroff MM, Haacke EM. Principles, techniques, and applications of T2*-based MR imaging and its special applications. Radiographics. 2009;29:1433–50.
