基于轻量化注意力残差网络的食管鳞癌识别方法

Esophageal Squamous Cell Carcinoma Recognition Based on Lightweight Residual Networks with an Attention Mechanism

食管鳞癌(ESCC)是我国常见的消化道恶性肿瘤之一。临床上,窄带成像联合放大内镜(NBI-ME)能够显示出食管粘膜层的微血管形态变化,是诊断ESCC的重要手段。针对ESCC识别模型难以兼顾识别准确率和推理效率的问题,提出一种融合注意力机制的轻量化残差网络(CALite-ResNet)对食管NBI-ME图像进行分类。从多家医院采集到206例患者共11468张NBI-ME图像作为本研究数据集。实验结果表明,ESCC识别的准确率和敏感度分别在图像级别达到96.39%和95.70%,在病人级别达到95.70%和94.62%,单张食管图像的平均预测时间为16.42 ms。因此,CALiteResNet模型对ESCC具有较高的识别准确率和较快的推理效率,能够为ESCC的临床辅助诊断提供有效帮助,具备一定的临床意义与应用价值。

Esophageal squamous cell carcinoma(ESCC)is one of the most common malignant digestive tract tumors in China.Clinically,narrowband imaging combined with magnifying endoscopy(NBIME)can be used to investigate the morphological changes of microvessels in the esophageal mucosa and serves as an important means of diagnosing ESCC.To solve the ESCC recognition model’s difficulty in considering both the recognition accuracy and reasoning efficiency,a lightweight residual network(CALiteResNet)with an integrated attention mechanism is proposed to classify esophageal NBIME images.The dataset for this study comprises 11468 NBIME images of 206 patients collected from multiple hospitals.The experimental results show that the accuracy and sensitivity of the ESCC recognition is 96.39%and 95.70%at the image level,and 95.70%and 94.62%at the patient level,respectively,and the average prediction time of a single esophageal image is 16.42 ms.Therefore,the CALiteResNet model has a higher recognition accuracy and faster reasoning efficiency for ESCC recognition,and a certain clinical significance and application value,thereby making it effective for use in the auxiliary clinical diagnosis of ESCC.