Management of Cerebral Aneurysm Descriptors based on an Automatic Ostium Extraction
M. Meuschke, T. Günther, R. Wickenhöfer, M. Gross, B. Preim, K. Lawonn
IEEE Computer Graphics and Applications, IEEE, vol. 38, no. 3, 2018, pp. 58-72
Abstract
We present a framework to manage cerebral aneurysm data, including morphological descriptors calculated from an automatically extracted ostium. Aneurysms bear the risk of rupture. Rupture risk evaluation is based on morphological descriptors, which in turn rely on a stable ostium detection–both are manually extracted by physicians. This time-consuming and error-prone process is incompatible with the high clinical workload and thus automatic solutions are desirable. Due to anatomical diversity, automatically generated results do not always reach manual results. Thus, automatic solutions should be able to incorporate expert knowledge. This raises several questions: How can expert knowledge be integrated? How should meta data be defined so that different clinicians prepare data consistently? Which interaction techniques are needed to explore large amounts of datasets? This paper explores these questions in collaboration with physicians. For adaption of visualization research in clinical routine, the integration of expert knowledge and consistent data documentation are inevitable.
