Houdeville Charles, Souchaud Marc, Leenhardt Romain, Goltstein Lia Cmj, Velut Guillaume, Beaumont Hanneke, Dray Xavier, Histace Aymeric
Sorbonne University, Center for Digestive Endoscopy, Saint-Antoine Hospital, APHP, 75012 Paris, France; Équipes Traitement de l'Information et Systèmes, ETIS UMR 8051, CY Paris Cergy University, ENSEA, CNRS, 95000 Cergy, France.
Équipes Traitement de l'Information et Systèmes, ETIS UMR 8051, CY Paris Cergy University, ENSEA, CNRS, 95000 Cergy, France.
Clin Res Hepatol Gastroenterol. 2025 Jan;49(1):102509. doi: 10.1016/j.clinre.2024.102509. Epub 2024 Nov 30.
Deep learning (DL) algorithms demonstrate excellent diagnostic performance for the detection of vascular lesions via small bowel (SB) capsule endoscopy (CE), including vascular abnormalities with high (P2), intermediate (P1) or low (P0) bleeding potential, while dramatically decreasing the reading time. We aimed to improve the performance of a DL algorithm by characterizing vascular abnormalities using a machine learning (ML) classifier, and selecting the most relevant images for insertion into reports.
A training dataset of 75 SB CE videos was created, containing 401 sequences of interest that encompassed 1,525 images of various vascular lesions. Several image classification algorithms were tested, to discriminate "typical angiodysplasia" (P2/P1) and "other vascular lesion" (P0) and to select the most relevant image within sequences with repetitive images. The performances of the best-fitting algorithms were subsequently assessed on an independent test dataset of 73 full-length SB CE video recordings.
Following DL detection, a random forest (RF) method demonstrated a specificity of 91.1 %, an area under the receiving operating characteristic curve of 0.873, and an accuracy of 84.2 % for discriminating P2/P1 from P0 lesions while allowing an 83.2 % reduction in the number of reported images. In the independent testing database, after RF was applied, the output number decreased by 91.6 %, from 216 (IQR 108-432) to 12 (IQR 5-33). The RF algorithm achieved 98.0 % agreement with initial, conventional (human) reporting. Following DL detection, the RF method allowed better characterization and accurate selection of images of relevant (P2/P1) SB vascular abnormalities for CE reporting without impairing diagnostic accuracy. These findings pave the way for automated SB CE reporting.
深度学习（DL）算法在通过小肠（SB）胶囊内镜（CE）检测血管病变方面表现出卓越的诊断性能，可检测出具有高（P2）、中（P1）或低（P0）出血风险的血管异常，同时显著减少阅片时间。我们旨在通过使用机器学习（ML）分类器对血管异常进行特征描述，并选择最相关的图像插入报告，以提高DL算法的性能。
创建了一个包含75个SB CE视频的训练数据集，其中包含401个感兴趣序列，涵盖1525张各种血管病变的图像。测试了几种图像分类算法，以区分“典型血管发育异常”（P2/P1）和“其他血管病变”（P0），并在具有重复图像的序列中选择最相关的图像。随后在一个包含73个全长SB CE视频记录的独立测试数据集上评估最佳拟合算法的性能。
在DL检测之后，随机森林（RF）方法在区分P2/P1与P0病变时，特异性为91.1%，接受者操作特征曲线下面积为0.873，准确率为84.2%，同时报告的图像数量减少了83.2%。在独立测试数据库中，应用RF后，输出数量从216（四分位间距108 - 432）减少到12（四分位间距5 - 33），减少了91.6%。RF算法与最初的传统（人工）报告达成了98.0%的一致性。在DL检测之后，RF方法能够更好地对SB相关（P2/P1）血管异常的图像进行特征描述和准确选择，用于CE报告，且不影响诊断准确性。这些发现为自动SB CE报告铺平了道路。
