DALSA: Domain Adaptation for Supervised Learning from Sparsely Annotated MR Images

The author proposes DALSA as a method to correct sampling selection errors introduced by sparse annotations, facilitating efficient tumor segmentation without sacrificing accuracy.