Computer-aided diagnosis of pituitary microadenoma on dynamic contrast-enhanced MRI based on spatio-temporal features

Te Guo, Jixin Luan, Jingyuan Gao, Bing Liu, Tianyu Shen, Hongwei Yu, Guolin Ma, Kunfeng Wang

Abstract: Highlights•Multiscale feature selection module to address the loss of feature information.•The Dual-path Semantic Segmentation Module solves the problem of precision loss.•Reuse the underlying information to overcome difficult detection of small targets.•Improving accuracy compared to U-NET series methods on a private dataset.