Post-mortem magnetic resonance imaging in fetuses, newborns and children

• Main Author: Thayyil, S. S.
• Published: University College London (University of London) 2010
• Subjects: 618.92
• Online Access: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.565059

My thesis explores the feasibility and utility of whole body post-mortem magnetic resonance (MR) imaging as an alternative for conventional autopsy in fetuses, newborns and children. The thesis starts with a systematic review of the existing literature on post-mortem MR imaging to identify the knowledge gaps. This is followed by the development of an effective recruitment model and a comparative study on the accuracy of less invasive autopsy by post-mortem MR imaging with conventional autopsy in 200 fetuses, newborns and children. The cause of death was accurately identified in more than 90% of cases by less invasive autopsy, following a hospital death, unexplained stillbirth or an unexpected death under HM Coronial investigation. Post-mortem MR imaging of the brain had a very high negative predictive value for excluding major neuropathological lesions; opening of the head can be avoided if post-mortem MR imaging of the brain is normal. High-resolution, 3D post-mortem cardiac MR imaging accurately detected structural heart diseases in larger fetuses, newborns and children. However, the accuracy of post-mortem lung MR imaging was poor; renal lesions required histological examination for definitive diagnosis. Furthermore, post-mortem MR imaging cannot differentiate between the normal death process and ante-mortem hypoxic brain injury due to the changes in T1 and T2 relaxometry values occurring after death. The diagnostic utility and image quality at 1.5 Tesla MR imaging was poor in smaller fetuses, however high field MR imaging at 9.4 Tesla provided satisfactory MR images in this sub group. In addition, visceral organ weights were accurately estimated from post-mortem MR data sets and anatomical models of these organs reconstructed by rapid prototyping. My thesis concludes by demonstrating the proof of principle of MR guided percutaneous biopsy in a piglet model. In summary, less invasive autopsy by post-mortem MR imaging may be a satisfactory alternative to conventional autopsy; however accurate methods of percutaneous tissue sampling need to be developed and validated for adequate histological examination of visceral organs, particularly lungs, kidneys and heart.