R. Stidham, Binu Enchakalody, Stewart C Wang, Grace L Su, Brian Ross, Mahmoud Al-Hawary, A. Wasnik
{"title":"Artificial Intelligence for Quantifying Cumulative Small Bowel Disease Severity on CT-Enterography in Crohn's Disease.","authors":"R. Stidham, Binu Enchakalody, Stewart C Wang, Grace L Su, Brian Ross, Mahmoud Al-Hawary, A. Wasnik","doi":"10.14309/ajg.0000000000002828","DOIUrl":null,"url":null,"abstract":"OBJECTIVE\nAssessing the cumulative degree of bowel injury in ileal Crohn's disease (CD) is difficult. We aimed to develop machine learning (ML) methodologies for automated estimation of cumulative ileal injury on CT-enterography (CTE) to help predict future bowel surgery.\n\n\nMETHODS\nAdults with ileal CD using biologic therapy at a tertiary care center underwent ML analysis of CTE scans. Two fellowship trained radiologists graded bowel injury severity at granular spatial increments along the ileum (1cm), called mini-segments. ML segmentation methods were trained on radiologist grading with predicted severity then spatially mapped to the ileum. Cumulative injury was calculated as the sum (S-CIDSS) and mean of severity grades along the ileum. Multivariate models of future small bowel resection were compared cumulative ileum injury metrics and traditional bowel measures, adjusting for laboratory values, medications, and prior surgery at the time of CTE.\n\n\nRESULTS\nIn 229 CTEs, 8424 mini-segments underwent analysis. Agreement between ML and radiologists injury grading was strong (κ=0.80, 95%CI 0.79-0.81) and similar to inter-radiologist agreement (κ=0.87, 95%CI 0.85-0.88). S-CIDSS (46.6 vs. 30.4, P=0.0007) and mean cumulative injury grade scores (1.80 vs. 1.42, P<0.0001) were greater in CD biologic users that went to future surgery. Models using cumulative spatial metrics (AUC=0.76) outperformed models using conventional bowel measures, laboratory values, and medical history (AUC=0.62) for predicting future surgery in biologic users.\n\n\nCONCLUSION\nAutomated cumulative ileal injury scores show promise for improving prediction of outcomes in small bowel CD. Beyond replicating expert judgement, spatial enterography analysis can augment the personalization of bowel assessment in CD.","PeriodicalId":507623,"journal":{"name":"The American Journal of Gastroenterology","volume":null,"pages":null},"PeriodicalIF":0.0000,"publicationDate":"2024-04-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"The American Journal of Gastroenterology","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.14309/ajg.0000000000002828","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0
Abstract
OBJECTIVE
Assessing the cumulative degree of bowel injury in ileal Crohn's disease (CD) is difficult. We aimed to develop machine learning (ML) methodologies for automated estimation of cumulative ileal injury on CT-enterography (CTE) to help predict future bowel surgery.
METHODS
Adults with ileal CD using biologic therapy at a tertiary care center underwent ML analysis of CTE scans. Two fellowship trained radiologists graded bowel injury severity at granular spatial increments along the ileum (1cm), called mini-segments. ML segmentation methods were trained on radiologist grading with predicted severity then spatially mapped to the ileum. Cumulative injury was calculated as the sum (S-CIDSS) and mean of severity grades along the ileum. Multivariate models of future small bowel resection were compared cumulative ileum injury metrics and traditional bowel measures, adjusting for laboratory values, medications, and prior surgery at the time of CTE.
RESULTS
In 229 CTEs, 8424 mini-segments underwent analysis. Agreement between ML and radiologists injury grading was strong (κ=0.80, 95%CI 0.79-0.81) and similar to inter-radiologist agreement (κ=0.87, 95%CI 0.85-0.88). S-CIDSS (46.6 vs. 30.4, P=0.0007) and mean cumulative injury grade scores (1.80 vs. 1.42, P<0.0001) were greater in CD biologic users that went to future surgery. Models using cumulative spatial metrics (AUC=0.76) outperformed models using conventional bowel measures, laboratory values, and medical history (AUC=0.62) for predicting future surgery in biologic users.
CONCLUSION
Automated cumulative ileal injury scores show promise for improving prediction of outcomes in small bowel CD. Beyond replicating expert judgement, spatial enterography analysis can augment the personalization of bowel assessment in CD.
目的评估回肠克罗恩病(CD)肠道损伤的累积程度非常困难。我们的目标是开发机器学习(ML)方法,自动估算 CT 肠道造影(CTE)的回肠累积损伤程度,帮助预测未来的肠道手术。方法在一家三级医疗中心接受生物治疗的回肠克罗恩病成人患者接受了 CTE 扫描的 ML 分析。两名接受过研究培训的放射科医生按照回肠(1 厘米)的颗粒空间增量对肠道损伤严重程度进行了分级,这些空间增量被称为小段。ML 分段方法是根据放射科医生的分级和预测的严重程度进行训练,然后在空间上映射到回肠。累积损伤以回肠严重程度分级的总和(S-CIDSS)和平均值计算。对未来小肠切除术的多变量模型进行了比较,比较了累积回肠损伤指标和传统的肠道测量指标,并对实验室值、药物和 CTE 时的既往手术进行了调整。ML与放射科医生损伤分级之间的一致性很强(κ=0.80,95%CI 0.79-0.81),与放射科医生之间的一致性相似(κ=0.87,95%CI 0.85-0.88)。在今后接受手术的 CD 生物制剂使用者中,S-CIDSS(46.6 vs. 30.4,P=0.0007)和平均累积损伤等级评分(1.80 vs. 1.42,P<0.0001)更高。在预测生物制剂使用者未来手术方面,使用累积空间指标的模型(AUC=0.76)优于使用传统肠道测量、实验室值和病史的模型(AUC=0.62)。除了复制专家的判断外,空间肠造影分析还能增强 CD 肠道评估的个性化。