Journal of Crohn's & colitis · 2026
AI-based quantification of inflammatory extent for relapse prediction in ulcerative colitis: a prospective cohort study.
СтатьяНазвание и аннотация: Перевод в очереди
Обработка статьи · этапы, ошибки и управление
Загрузка состояния…
Ошибки и ограничения
История · было → стало
Состояние обновляется каждые 5 секунд, пока блок открыт. Процент показывается только там, где известен объём работы. Пауза останавливает продолжение этапа; текущий запрос к источнику может завершиться.
Maeda Yasuharu, Kudo Shin-Ei, Ogata Noriyuki, Takenaka Kento, Takabayashi Kaoru, Kuroki Takanori, Kawabata Yurie, Okumura Taishi, Sakurai Tatsuya, Kouyama Yuta, Ichimasa Katsuro, Hayashi Takemasa, Baba Toshiyuki, Ogata Haruhiko, Ohtsuka Kazuo, Mori Yuichi, Iacucci Marietta, Misawa Masashi
Аннотация
BACKGROUND AND AIMS: Endoscopic remission is a key therapeutic goal in ulcerative colitis (UC); however, conventional indices focus on peak severity without accounting for the spatial extent of inflammation. We developed a deep learning-based score, Quantitative Ulcerative Colitis Assessment using Deep Learning (QUAD), and evaluated whether incorporating inflammatory extent improves relapse prediction in patients with UC in clinical remission. To our knowledge, no studies have examined whether artificial intelligence (AI)-derived assessment of inflammatory extent predicts clinical relapse.
METHODS: This prospective cohort study evaluated relapse prediction over 24 months in patients with UC in clinical remission. The QUAD model assigns a score of 0-3 to each image quadrant, yielding a total score of 0-12. The model was trained on 84 743 images from 998 patients with UC across three centers. Still image-based validation and automated full-length video analysis were conducted to assess the impact of inflammatory extent on relapse prediction.
RESULTS: Clinical relapse occurred in 19.4% of patients with QUAD ≥ 4 compared with 5.2% of those with QUAD < 4 (P = .01), with an area under the curve (AUC) of 0.67 (95% confidence interval [CI]: 0.57-0.76). Notably, automated video-based analysis showed that distal inflammatory burden yielded the highest predictive performance, with analysis of the distal 10% segment achieving an AUC of 0.73 (95% CI: 0.62-0.85), compared with whole-colon assessment (AUC, 0.62; 95% CI: 0.48-0.76) and still image-based evaluation.
CONCLUSION: AI-augmented assessment integrating inflammatory severity and extent may provide complementary prognostic information beyond severity-based evaluation.
METHODS: This prospective cohort study evaluated relapse prediction over 24 months in patients with UC in clinical remission. The QUAD model assigns a score of 0-3 to each image quadrant, yielding a total score of 0-12. The model was trained on 84 743 images from 998 patients with UC across three centers. Still image-based validation and automated full-length video analysis were conducted to assess the impact of inflammatory extent on relapse prediction.
RESULTS: Clinical relapse occurred in 19.4% of patients with QUAD ≥ 4 compared with 5.2% of those with QUAD < 4 (P = .01), with an area under the curve (AUC) of 0.67 (95% confidence interval [CI]: 0.57-0.76). Notably, automated video-based analysis showed that distal inflammatory burden yielded the highest predictive performance, with analysis of the distal 10% segment achieving an AUC of 0.73 (95% CI: 0.62-0.85), compared with whole-colon assessment (AUC, 0.62; 95% CI: 0.48-0.76) and still image-based evaluation.
CONCLUSION: AI-augmented assessment integrating inflammatory severity and extent may provide complementary prognostic information beyond severity-based evaluation.