TransCC: Transformer Network for Coronary Artery CCTA Segmentation

TL;DR

TransCC is a novel deep learning framework combining Transformer and CNN techniques to improve coronary artery segmentation in CCTA images, addressing local structure damage and enhancing global-local feature integration.

Contribution

The paper introduces TransCC, a new model integrating a Feature Interaction Extraction module and a Multilayer Enhanced Perceptron to improve segmentation accuracy in CCTA images.

Findings

TransCC achieves an average Dice coefficient of 0.730.

TransCC attains an average IoU of 0.582.

TransCC outperforms existing segmentation methods.

Abstract

The accurate segmentation of Coronary Computed Tomography Angiography (CCTA) images holds substantial clinical value for the early detection and treatment of Coronary Heart Disease (CHD). The Transformer, utilizing a self-attention mechanism, has demonstrated commendable performance in the realm of medical image processing. However, challenges persist in coronary segmentation tasks due to (1) the damage to target local structures caused by fixed-size image patch embedding, and (2) the critical role of both global and local features in medical image segmentation tasks.To address these challenges, we propose a deep learning framework, TransCC, that effectively amalgamates the Transformer and convolutional neural networks for CCTA segmentation. Firstly, we introduce a Feature Interaction Extraction (FIE) module designed to capture the characteristics of image patches, thereby circumventing the loss of semantic information inherent in the original method. Secondly, we devise a Multilayer Enhanced Perceptron (MEP) to augment attention to local information within spatial dimensions, serving as a complement to the self-attention mechanism. Experimental results indicate that TransCC outperforms existing methods in segmentation performance, boasting an average Dice coefficient of 0.730 and an average Intersection over Union (IoU) of 0.582. These results underscore the effectiveness of TransCC in CCTA image segmentation.