Tang Suigu, Yu Xiaoyuan, Cheang Chak Fong, Ji Xiaoyu, Yu Hon Ho, Choi I Cheong
Faculty of Innovation Engineering-School of Computer Science and Engineering, Macau University of Science and Technology, Avenida Wai Long, Taipa, Macau SAR.
Faculty of Innovation Engineering-School of Computer Science and Engineering, Macau University of Science and Technology, Avenida Wai Long, Taipa, Macau SAR.
Comput Methods Programs Biomed. 2023 Apr;231:107399. doi: 10.1016/j.cmpb.2023.107399. Epub 2023 Feb 8.
A deep learning-based intelligent diagnosis system can significantly reduce the burden of endoscopists in the daily analysis of esophageal lesions. Considering the need to add new tasks in the diagnosis system, a deep learning model that can train a series of tasks incrementally using endoscopic images is essential for identifying the types and regions of esophageal lesions.
In this paper, we proposed a continual learning-based esophageal lesion network (CLELNet), in which a convolutional autoencoder was designed to extract representation features of endoscopic images among different esophageal lesions. The proposed CLELNet consists of shared layers and task-specific layers. Shared layers are used to extract common features among different lesions while task-specific layers can complete different tasks. The first two tasks trained by the CLELNet are the classification (task 1) and the segmentation (task 2). We collected a dataset of esophageal endoscopic images from Macau Kiang Wu Hospital for training and testing the CLELNet.
The experimental results showed that the classification accuracy of task 1 was 95.96%, and the Intersection Over Union and the Dice Similarity Coefficient of task 2 were 65.66% and 78.08%, respectively.
The proposed CLELNet can realize task-incremental learning without forgetting the previous tasks and thus become a useful computer-aided diagnosis system in esophageal lesions analysis.
基于深度学习的智能诊断系统可显著减轻内镜医师在日常食管病变分析中的负担。考虑到诊断系统中需要添加新任务，一个能够使用内镜图像逐步训练一系列任务的深度学习模型对于识别食管病变的类型和区域至关重要。
在本文中，我们提出了一种基于持续学习的食管病变网络（CLELNet），其中设计了一个卷积自动编码器来提取不同食管病变内镜图像的表征特征。所提出的CLELNet由共享层和特定任务层组成。共享层用于提取不同病变之间的共同特征，而特定任务层可以完成不同的任务。CLELNet训练的前两个任务是分类（任务1）和分割（任务2）。我们从澳门镜湖医院收集了食管内镜图像数据集用于训练和测试CLELNet。
实验结果表明，任务1的分类准确率为95.96%，任务2的交并比和骰子相似系数分别为65.66%和78.08%。
所提出的CLELNet可以在不遗忘先前任务的情况下实现任务增量学习，从而成为食管病变分析中一个有用的计算机辅助诊断系统。
