Deep Learning-Based Identification of Ischemic Regions in Native Head CT Scans

Sammanfattning: Stroke is one of the major causes of death and disability worldwide. Fast diagnosis is of critical importance for stroke treatment. In clinical routine, a non-contrast CT (NCCT) is typically acquired immediately to determine whether the stroke is ischemic or hemorrhagic and plan therapy accordingly. In case of ischemia, early signs of infarction may appear due to increased water uptake. These signs may be subtle, especially if observed only shortly after symptom onset, but hold the potential to provide a crucial first assessment of the location and extent of the infarction. In this paper, we train a deep neural network to predict the infarct core from NCCT in an image-to-image fashion. To facilitate exploitation of anatomic correspondences, learning is carried out in the standardized coordinate system of a brain atlas to which all images are deformably registered. Apart from binary infarct core masks, perfusion maps such as cerebral blood volume and flow are employed as additional training targets to enrich the physiologic information available to the model. The method is evaluated using cross validation on the training data set consisting of 141 cases. For validation, we measure the overlap with the ground truth masks, the localisation performance and the agreement with both manual and automatic assessment of affected ASPECTS regions. It is shown that the additional targets improve the results signficantly, achieving an area-under-curve of 0.835 when compared with the automated assessment in ASPECTS region classification and providing a distance of 0 mm between the prediction maximum and the indicated stroke infarct core in the majority of severe strokes with an infarct core volume greater than 70 ml.