ECA-UNet for coronary artery segmentation and three-dimensional reconstruction

Article Computer Science, Artificial Intelligence

Automatic coronary artery segmentation algorithm based on deep learning and digital image processing

Fangzheng Tian et al.

Summary: The study proposes a novel CCTA image segmentation framework that combines deep learning and digital image processing algorithms, utilizing V-Net and patch-based cascaded V-shaped module to achieve good results in coronary artery segmentation.

APPLIED INTELLIGENCE (2021)

Add to Collection

Article Physiology

ImageTBAD: A 3D Computed Tomography Angiography Image Dataset for Automatic Segmentation of Type-B Aortic Dissection

Zeyang Yao et al.

Summary: Type-B Aortic Dissection (TBAD) is a serious cardiovascular event with increasing incidence and poor prognosis. Existing research focuses on true lumen (TL) and false lumen (FL) without considering false lumen thrombus (FLT). This study introduces the ImageTBAD dataset and a baseline method for automatic segmentation, highlighting the challenge of FLT segmentation and the need for improvement.

FRONTIERS IN PHYSIOLOGY (2021)

Add to Collection

Article Energy & Fuels

Improved U-Net based insulator image segmentation method based on attention mechanism

Han Gujing et al.

Summary: This paper proposes an insulator segmentation method based on improved U-Net, which embeds the attention mechanism ECA-Net to enhance the model's ability in extracting semantic features and improving the accuracy of insulator detection. The experimental results demonstrate that the proposed method achieves an average overlap IOU of 96.8%, enabling more accurate segmentation of different types of insulators in complex backgrounds.

ENERGY REPORTS (2021)

Add to Collection

Article Medicine, General & Internal