Thus, the team developed DeepWMH, an open-source, annotation-free WMH lesion segmentation tool targeted for removing the need for human annotation, while achieving high segmentation accuracy.