15283-B-Blokker

174 Chapter 9 The MIA brain biopsies usually did not interfere with the execution and interpretation of the conventional brain autopsy. However, in one case the brain autopsy detected a focal intracerebral bleeding that was probably caused by the biopsies, for it was, even in retrospect, not present on the imaging. We assume that, if MIA were to be clinically implemented in the future, and histological examination of the brain is permitted by next-of-kin, either MIA-biopsies or conventional brain autopsy but not both will be performed. In that situation, the images could be used for either obtaining the brain biopsies or performing the brain autopsy, but the biopsies will no longer cause artifactual findings at brain autopsy. Small or subtle pathologies Another important difficultywithMIA is thedetectionof small lesions, suchas (recurrent) tumor processes and metastases 80 , or to radiologically distinguish inflammation processes from, for example, hematological malignancies. For such pathologies it is helpful to compare post-mortem images with recent pre-mortem scans whenever available and to obtain tissue samples for further examination. Patients who were known to have cancer, for example, have probably been scanned for follow-up, thus recurrent tumor growth or metastases are likely to be known at time of death. When small suspicious lesions are identified at the post-mortem images, it may be difficult to obtain representative biopsies, especially if they measure less than 1 cm in diameter, are connected to soft tissue structures that move along with the biopsy needle, and are poorly visible during the biopsy procedure, due to metal artifacts of the biopsy needle. In our cohort, we were unable to identify or confirm three squamous cell carcinomas of the respiratory tract with MIA. One recurrent squamous cell carcinoma in the area of mouth, trachea, larynx and lung was missed on imaging (perceptual error) although visible on pre-mortem images, and two squamous cell carcinomas in the lung were missed by biopsy (sampling error). TypicallyMIAwas unable to answer clinical questions that concerned subtle pathological changes not visible with the applied imaging techniques. For example, the status of surgical sutures could not be realisably imaged without the use of contrast (2 cases). Along the same lines, it was impossible to identify intoxication with certainty as the cause of death without toxicological examination. 52