AsT: An Asymmetric-Sensitive Transformer for Osteonecrosis of the Femoral Head Detection (Student Abstract)

Early diagnosis of osteonecrosis the femoral head (ONFH) can inhibit progression and improve preservation. The radiograph difference between early ONFH healthy ones is not apparent to naked eye. It also hard produce a large dataset train classification model. In this paper, we propose Asymmetric-Sensitive Transformer (AsT) capture uneven development bilateral enable robust detection. Our detection realized using self-attention mechanism regions while conferring sensitivity by attention-shared transformer. real-world experiment studies show that AsT achieves best performance AUC 0.9313 in find out misdiagnosis cases firmly.