University College London Hospitals, London, United Kingdom.
School of Biomedical Engineering & Imaging Sciences, Kings College London, London, United Kingdom.
Gastrointest Endosc. 2021 Aug;94(2):273-281. doi: 10.1016/j.gie.2021.01.043. Epub 2021 Feb 5.
Intrapapillary capillary loops (IPCLs) are microvascular structures that correlate with the invasion depth of early squamous cell neoplasia and allow accurate prediction of histology. Artificial intelligence may improve human recognition of IPCL patterns and prediction of histology to allow prompt access to endoscopic therapy for early squamous cell neoplasia where appropriate.
One hundred fifteen patients were recruited at 2 academic Taiwanese hospitals. Magnification endoscopy narrow-band imaging videos of squamous mucosa were labeled as dysplastic or normal according to their histology, and IPCL patterns were classified by consensus of 3 experienced clinicians. A convolutional neural network (CNN) was trained to classify IPCLs, using 67,742 high-quality magnification endoscopy narrow-band images by 5-fold cross validation. Performance measures were calculated to give an average F1 score, accuracy, sensitivity, and specificity. A panel of 5 Asian and 4 European experts predicted the histology of a random selection of 158 images using the Japanese Endoscopic Society IPCL classification; accuracy, sensitivity, specificity, positive and negative predictive values were calculated.
Expert European Union (EU) and Asian endoscopists attained F1 scores (a measure of binary classification accuracy) of 97.0% and 98%, respectively. Sensitivity and accuracy of the EU and Asian clinicians were 97%, 98% and 96.9%, 97.1%, respectively. The CNN average F1 score was 94%, sensitivity 93.7%, and accuracy 91.7%. Our CNN operates at video rate and generates class activation maps that can be used to visually validate CNN predictions.
We report a clinically interpretable CNN developed to predict histology based on IPCL patterns, in real time, using the largest reported dataset of images for this purpose. Our CNN achieved diagnostic performance comparable with an expert panel of endoscopists.
内乳头毛细血管袢（IPCLs）是与早期鳞状细胞癌浸润深度相关的微血管结构，能够准确预测组织学。人工智能可能会提高人类对 IPCL 模式的识别能力，以及对组织学的预测能力，从而在适当的情况下，为早期鳞状细胞癌提供及时的内镜治疗。
在台湾的 2 家学术医院招募了 115 名患者。根据组织学将鳞状黏膜的放大内镜窄带成像视频标记为异型增生或正常，并由 3 名经验丰富的临床医生共识分类 IPCL 模式。通过 5 倍交叉验证，使用 67742 张高质量的放大内镜窄带图像对卷积神经网络（CNN）进行训练以分类 IPCL。计算性能指标以给出平均 F1 评分、准确性、敏感性和特异性。一个由 5 名亚洲专家和 4 名欧洲专家组成的小组使用日本内镜学会 IPCL 分类法预测了 158 张随机选择图像的组织学；计算了准确性、敏感性、特异性、阳性和阴性预测值。
欧洲专家和亚洲内镜医生的 F1 评分（衡量二分类准确性的指标）分别为 97.0%和 98%。欧洲和亚洲临床医生的敏感性和准确性分别为 97%、98%和 96.9%、97.1%。CNN 的平均 F1 得分为 94%，敏感性为 93.7%，准确性为 91.7%。我们的 CNN 以视频速度运行，并生成类激活图，可用于可视化验证 CNN 预测。
我们报告了一种临床可解释的 CNN，该网络能够根据 IPCL 模式实时预测组织学，使用为此目的报告的最大图像数据集。我们的 CNN 实现了与内镜专家小组相当的诊断性能。
