ADSeg: A flap-attention-based deep learning approach for aortic dissection segmentation

Accurate and rapid segmentation of the lumen in an aortic dissection (AD) is an important prerequisite for risk evaluation and medical planning for patients with this serious condition. Although some recent studies have pioneered technical advances for the challenging AD segmentation task, they generally neglect the intimal flap structure that separates the true and false lumens. Identification and segmentation of the intimal flap may simplify AD segmentation, and the incorporation of long-distance z axis information interaction along the curved aorta may improve segmentation accuracy. This study proposes a flap attention module that focuses on key flap voxels and performs operations with long-distance attention. In addition, a pragmatic cascaded network structure with feature reuse and a two-step training strategy are presented to fully exploit network representation power. The proposed ADSeg method was evaluated on a multicenter dataset of 108 cases, with or without thrombus; ADSeg outperformed previous state-of-the-art methods by a significant margin and was robust against center variation. •A novel flap attention module is proposed for segmenting type B aortic dissection•A cascaded network structure with feature reuse and a two-step strategy is presented•Evaluation was performed on a multicenter dataset including with or without thrombus•The proposed method outperforms previous state-of-the-art deep learning methods An aortic dissection (AD) is a serious condition in which the inner layer of the aorta tears, causing the inner and middle tissue layers to split as blood surges through the tear. Image processing to segment and visualize the anatomy of an AD is essential for disease diagnosis, surgical planning, and postoperative follow-up. In this study, we propose a deep learning method to improve the accuracy of image processing when segmenting the true and false lumen of an AD. When compared with the current optimal methods, our method more accurately discerns the anatomic features of AD, which may increase the likelihood for successful surgery and reduction of postoperative complications. Our overarching goal is to develop an intelligent platform for accurate diagnosis of AD, individualized surgical planning, and prognosis prediction, which will improve the survival of patients with AD. Given the high mortality rate of AD, any improvement in diagnosis and treatment efficiency could have substantial benefits for the healthcare system and patient well-b