Automatic Hippocampal Multimodal Assessment for Studies of Stroke and Small Vessel Disease

We propose a framework for assessing the hippocampi on stroke patients and studies of small vessel disease, where sclerosis, perivascular spaces and infarcts on this structure are common. It includes hippocampal and cavity segmentations, hippocampal shape modelling, feature characterisation and statistical analyses, all which have been particularly developed for assessing extreme abnormalities on this small brain structure on clinical MRI datasets. The hippocampal segmentation uses FSL tools. We apply this approach to 48 datasets from a study of mild stroke and assess its relevance on a larger dataset of 189 stroke patients. Using shape similarity metrics we show that the hippocampal shape models generated by our method are accurate on this dataset. We estimate the prevalence and distribution of the features analysed on the samples to discuss the usefulness of our approach.