Self-supervised learning for stroke lesion segmentation on CT: a new pretext task for neuroimaging.

With the advancement of technology, the availability of medical data has increased, enabling the training of deep learning models for automatic analysis of medical images. However, annotating this data can be costly. To address this challenge, self-s…

Preview only. Read the full abstract at the source