Department of Gastroenterology, International University of Health and Welfare Ichikawa Hospital, 6-1-14, Konodai, Ichikawa-Shi, Chiba, 272-0827, Japan.
Department of Systems and Control Engineering, School of Engineering, Tokyo Institute of Technology, Tokyo, Japan.
Gastric Cancer. 2024 May;27(3):539-547. doi: 10.1007/s10120-024-01469-7. Epub 2024 Jan 19.
Cycle-consistent generative adversarial network (CycleGAN) is a deep neural network model that performs image-to-image translations. We generated virtual indigo carmine (IC) chromoendoscopy images of gastric neoplasms using CycleGAN and compared their diagnostic performance with that of white light endoscopy (WLE).
WLE and IC images of 176 patients with gastric neoplasms who underwent endoscopic resection were obtained. We used 1,633 images (911 WLE and 722 IC) of 146 cases in the training dataset to develop virtual IC images using CycleGAN. The remaining 30 WLE images were translated into 30 virtual IC images using the trained CycleGAN and used for validation. The lesion borders were evaluated by 118 endoscopists from 22 institutions using the 60 paired virtual IC and WLE images. The lesion area concordance rate and successful whole-lesion diagnosis were compared.
The lesion area concordance rate based on the pathological diagnosis in virtual IC was lower than in WLE (44.1% vs. 48.5%, p < 0.01). The successful whole-lesion diagnosis was higher in the virtual IC than in WLE images; however, the difference was insignificant (28.2% vs. 26.4%, p = 0.11). Conversely, subgroup analyses revealed a significantly higher diagnosis in virtual IC than in WLE for depressed morphology (41.9% vs. 36.9%, p = 0.02), differentiated histology (27.6% vs. 24.8%, p = 0.02), smaller lesion size (42.3% vs. 38.3%, p = 0.01), and assessed by expert endoscopists (27.3% vs. 23.6%, p = 0.03).
The diagnostic ability of virtual IC was higher for some lesions, but not completely superior to that of WLE. Adjustments are required to improve the imaging system's performance.
循环一致生成对抗网络（CycleGAN）是一种深度神经网络模型，可执行图像到图像的转换。我们使用 CycleGAN 生成虚拟靛蓝胭脂红（IC）色素内镜胃肿瘤图像，并将其与白光内镜（WLE）的诊断性能进行比较。
获取 176 例接受内镜切除术的胃肿瘤患者的 WLE 和 IC 图像。我们使用来自 22 个机构的 118 名内镜医生，在训练数据集内的 146 例患者的 1633 张图像（911 张 WLE 和 722 张 IC）上，使用 CycleGAN 生成虚拟 IC 图像。剩余的 30 张 WLE 图像通过经过训练的 CycleGAN 转化为 30 张虚拟 IC 图像，用于验证。
基于病理诊断的虚拟 IC 病变面积一致性率低于 WLE（44.1%比 48.5%，p<0.01）。虚拟 IC 图像中整体病变的诊断成功率高于 WLE 图像，但差异无统计学意义（28.2%比 26.4%，p=0.11）。相反，亚组分析显示，在凹陷形态（41.9%比 36.9%，p=0.02）、分化组织学（27.6%比 24.8%，p=0.02）、较小病变大小（42.3%比 38.3%，p=0.01）以及专家内镜医生评估时（27.3%比 23.6%，p=0.03），虚拟 IC 的诊断能力均显著高于 WLE。
虚拟 IC 的诊断能力对于某些病变更高，但并不完全优于 WLE。需要进行调整以改善成像系统的性能。
