Cross-Shaped Windows Transformer with Self-supervised Pretraining for Clinically Significant Prostate Cancer Detection in Bi-parametric MRI

arXiv:2305.00385v2 Announce Type: replace-cross
Abstract: Biparametric magnetic resonance imaging (bpMRI) has demonstrated promising results in prostate cancer (PCa) detection using convolutional neural networks (CNNs). Recently, transformers have achieved competitive performance compared to CNNs in computer vision. Large scale transformers need abundant annotated data for training, which are difficult to obtain in medical imaging. Self-supervised learning (SSL) utilizes unlabeled data to generate meaningful semantic representations without the need for costly annotations, enhancing model performance on tasks with limited labeled data. We …

abstract annotated data arxiv cancer cancer detection cnns computer computer vision convolutional neural networks cs.cv data detection eess.iv imaging mri networks neural networks parametric performance pretraining results scale training transformer transformers type vision windows