Utilizing a novel Cross-Shaped Windows Transformer model with self-supervised pretraining significantly enhances clinically significant prostate cancer detection in MRI.
