Skip to content
DraconiticResearch
Search Sources API
203,203 studies31 sourcesUpdated 54 min ago
Peer-reviewedCC BYScientific Reports

Automated assessment of the ulcerative colitis endoscopic index of severity using a multi-task deep learning model

  • Bing Lv1
  • Qiang Zheng1
  • Xinxin Li1
  • Tao Tao2
  • Jingjing Wu3
  • Yanting Shi2
  1. 1School of Computer Science and Technology, Shandong University of Technology
  2. 2Department of Gastroenterology, Zibo Central Hospital
  3. 3Department of Internal Medicine, Zhangdian Maternal and Child Health Care Hospital
Read full textPDFDOICite

Abstract

Assessment of the Ulcerative Colitis Endoscopic Index of Severity (UCEIS) is limited by subjectivity and interobserver variability. We developed UC-MTLNet, a multi-task deep learning model to predict UCEIS descriptors, total score, endoscopic remission, and severity strata. This multicenter diagnostic study included 405 patients and 10,269 white-light endoscopic images from a development cohort, a prospectively collected internal test cohort, and an independent external test cohort. UC-MTLNet was evaluated at image and patient levels against a consensus reference standard. Descriptor-level QWK ranged from 0.9077 to 0.9337 in the internal test set and from 0.9142 to 0.9313 in the external test set. For total UCEIS, QWK was 0.9690 at image level and 0.9729 at patient level in the internal test set, with corresponding external values of 0.9646 and 0.9783. Image-level accuracies were 94.25%–95.66% for remission identification and 86.74%–89.21% for four-category severity stratification. Under standardized image-based scoring conditions, UC-MTLNet showed significantly higher exact-score accuracy and QWK than five experienced endoscopists across all settings. Compared with three single-task models, UC-MTLNet provided comparable performance with approximately one-third as many parameters and FLOPs and 2.78-fold faster inference. UC-MTLNet may assist standardized UCEIS assessment in research and central reading, pending prospective video-based and outcome-linked validation.

Related

  1. Deep learning-based assessment of ulcerative colitis activity from full-length endoscopic videos with spatial characterisation and histological correlation
    A. Bogush, A. Toskas, G. Ralli +11
    medRxivPreprintCC BY-NCFull textGenetics & genomicsmedRxiv
  2. Pitfalls and challenges in endoscopic assessment of ulcerative colitis.
    Sacha Mitchell, Chelssy Guerine Ingabire, Megan Oleksiw +2
    EndoscopyPeer-reviewedCC BYFull textGenetics & genomicsEurope PMC
  3. Bilirubin statistically accounts for the association between Enterobacterales lineage and Mayo endoscopic score in ulcerative colitis: a cross-sectional associational variable analysis.
    Zhengyu Ren, Junni Ma, Siyao Wang +4
    Frontiers in medicinePeer-reviewedCC BYFull textGenetics & genomicsEurope PMCPMC
  4. Diagnostic prediction of ulcerative colitis using simplified endoscopic features and blood count-derived indices: a multicenter study
    Yunlong Sun, Dongyu Cheng, Hairong Huang +6
    Frontiers in MedicinePeer-reviewedCC BYFull textGenetics & genomicsDOAJ
  5. Bowel Urgency Is More Strongly Associated with Biochemical, Endoscopic, and Histological Disease Activity than Other Clinical Symptoms in Ulcerative Colitis
    Ryohei Ogihara, Hiroki Kurumi, Koichiro Kawaguchi +2
    Internal MedicinePeer-reviewedCC BY-NC-NDFull textGenetics & genomicsJ-STAGE

Cite

BibTeXRIS
BibTeX
@article{lv2026automated,
  title = {{Automated assessment of the ulcerative colitis endoscopic index of severity using a multi-task deep learning model}},
  author = {Bing Lv and Qiang Zheng and Xinxin Li and Tao Tao and Jingjing Wu and Yanting Shi},
  journal = {Scientific Reports},
  year = {2026},
  month = jul,
  doi = {10.1038/s41598-026-64283-4},
  url = {https://doi.org/10.1038/s41598-026-64283-4},
  note = {CC BY},
}
RIS
TY  - JOUR
TI  - Automated assessment of the ulcerative colitis endoscopic index of severity using a multi-task deep learning model
AU  - Bing Lv
AU  - Qiang Zheng
AU  - Xinxin Li
AU  - Tao Tao
AU  - Jingjing Wu
AU  - Yanting Shi
PY  - 2026
DA  - 2026/07/01
JO  - Scientific Reports
DO  - 10.1038/s41598-026-64283-4
UR  - https://doi.org/10.1038/s41598-026-64283-4
AB  - Assessment of the Ulcerative Colitis Endoscopic Index of Severity (UCEIS) is limited by subjectivity and interobserver variability. We developed UC-MTLNet, a multi-task deep learning model to predict UCEIS descriptors, total score, endoscopic remission, and severity strata. This multicenter diagnostic study included 405 patients and 10,269 white-light endoscopic images from a development cohort, a prospectively collected internal test cohort, and an independent external test cohort. UC-MTLNet was evaluated at image and patient levels against a consensus reference standard. Descriptor-level QWK ranged from 0.9077 to 0.9337 in the internal test set and from 0.9142 to 0.9313 in the external test set. For total UCEIS, QWK was 0.9690 at image level and 0.9729 at patient level in the internal test set, with corresponding external values of 0.9646 and 0.9783. Image-level accuracies were 94.25%–95.66% for remission identification and 86.74%–89.21% for four-category severity stratification. Under standardized image-based scoring conditions, UC-MTLNet showed significantly higher exact-score accuracy and QWK than five experienced endoscopists across all settings. Compared with three single-task models, UC-MTLNet provided comparable performance with approximately one-third as many parameters and FLOPs and 2.78-fold faster inference. UC-MTLNet may assist standardized UCEIS assessment in research and central reading, pending prospective video-based and outcome-linked validation.
SN  - 2045-2322
ER  -