Currently submitted to: JMIR Medical Informatics
Date Submitted: Jan 31, 2026
Open Peer Review Period: Feb 20, 2026 - Apr 17, 2026
(currently open for review)
Warning: This is an author submission that is not peer-reviewed or edited. Preprints - unless they show as "accepted" - should not be relied on to guide clinical practice or health-related behavior and should not be reported in news media as established information.
Automated Lewis Score Calculation in Crohn’s Disease: A Multi-Task Artificial Intelligence Framework：Retrospective Study
ABSTRACT
Background:
Inflammatory Bowel Disease (IBD) is a chronic nonspecific intestinal inflammatory condition; accurate severity assessment is critical for clinical treatment decisions and prognosis. Current IBD evaluation relies primarily on endoscopic examinations and physician expertise, which are subjective and inconsistent.
Objective:
This study aimed to develop an automated deep learning-based scoring algorithm for objective quantitative assessment of IBD lesion severity to address the clinical challenge of subjective evaluation.
Methods:
A multi-stage deep learning architecture was employed for automatic IBD scoring: (1) an improved YOLO-V11 segmentation network (enhanced by attention mechanisms and multi-scale feature fusion) precisely identified edema and ulcer regions in intestinal endoscopic images; (2) a classification module based on YOLO-V11-derived lesion features recognized stenosis; (3) an LSTM lightweight normalization network integrated spatial and temporal lesion features to generate comprehensive IBD scores. Validation used endoscopic video data from 814 patients at a large medical center: 725 cases (4,400 annotated images) for edema/ulcer/stenosis recognition, and 89 cases for comprehensive scoring.
Results:
The model’s automatic scoring results showed a mean squared error (MSE) of 14.693 and a coefficient of determination (R²) of 0.82 compared with expert scoring. Key innovations included: (1) first combination of lesion recognition algorithms with image distribution frequency features for a multi-dimensional evaluation system; (2) development of a lightweight lesion recognition network suitable for clinical settings; (3) establishment of a large-scale annotated dataset encompassing various IBD subtypes.
Conclusions:
This automated scoring system improves the objectivity and repeatability of IBD severity assessment, providing a reliable tool for telemedicine and clinical trials. Future research will optimize the model’s performance in pediatric IBD and small bowel lesions.
Citation
Request queued. Please wait while the file is being generated. It may take some time.
Copyright
© The authors. All rights reserved. This is a privileged document currently under peer-review/community review (or an accepted/rejected manuscript). Authors have provided JMIR Publications with an exclusive license to publish this preprint on it's website for review and ahead-of-print citation purposes only. While the final peer-reviewed paper may be licensed under a cc-by license on publication, at this stage authors and publisher expressively prohibit redistribution of this draft paper other than for review purposes.