Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2025 Aug 13:18:17562848251364194.
doi: 10.1177/17562848251364194. eCollection 2025.

Artificial intelligence-based multimodal model for the identification of ulcerative colitis with concomitant cytomegalovirus colitis

Haozheng Liang  1 , Yuxuan Tian  2 , Gechong Ruan  1 , Xiaoyin Bai  1 , Wei Han  3 , Xiangling Fu  4 , Yuhang Wang  4 , Jialin Shi  5 , Yinghao Sun  1 , Ji Wu  6 , Chenyi Guo  6 , Hong Yang  7
Affiliations

Artificial intelligence-based multimodal model for the identification of ulcerative colitis with concomitant cytomegalovirus colitis

Haozheng Liang et al. Therap Adv Gastroenterol. .

Abstract

Background: Ulcerative colitis (UC), a chronic immune-mediated colon inflammation, impacts patients' quality of life. Immunosuppressive-treated UC patients are prone to opportunistic infections like cytomegalovirus (CMV) infection, which exacerbates UC, causes steroid resistance, and elevates surgery and mortality risks. Identifying CMV colitis from UC exacerbation is difficult due to overlapping symptoms and low biopsy detection rates.

Objectives: To develop an artificial intelligence (AI)-based multimodal model for early identification of UC with concomitant CMV colitis.

Design: This was a retrospective diagnostic study.

Methods: A total of 174 moderate to severe UC patients (87 with CMV colitis) from 2015 to 2023 in Peking Union Medical College Hospital were enrolled retrospectively. A total of 3345 colonoscopy images were collected. The dataset was split into training (70%) and testing (30%) sets. A multimodal dynamic affine transformation (DAFT) model integrating clinical biomarkers and endoscopic images was constructed, along with ResNet and SeNet models. Model performance was evaluated using accuracy, sensitivity, specificity, positive and negative predictive values from the confusion matrix.

Results: UC patients with CMV colitis had distinct clinical characteristics. The multimodal DAFT model outperformed ResNet and SeNet in distinguishing UC with CMV colitis, with higher accuracy (0.91), sensitivity (0.87), and specificity (0.93).

Conclusion: AI application offers a promising way to enhance early identification of UC with CMV colitis. The multimodal model combining clinical and endoscopic data can assist clinicians in accurate and timely diagnosis.

Keywords: artificial intelligence; cytomegalovirus colitis; deep learning; multimodal model; ulcerative colitis.

Plain language summary

Artificial intelligence based multimodal model for the identification of ulcerative colitis with concomitant cytomegalovirus colitis Background: Ulcerative colitis (UC), a chronic immune-mediated colon inflammation, impacts patients’ quality of life. Immunosuppressive-treated UC patients are prone to opportunistic infections like cytomegalovirus (CMV) infection, which exacerbates UC, causes steroid resistance, and elevates surgery and mortality risks. Identifying CMV colitis from UC exacerbation is difficult due to overlapping symptoms and low biopsy detection rates. Aim: To develop an AI-based multimodal model for early identification of UC with concomitant CMV colitis. Methods: 174 moderate to severe UC patients (87 with CMV colitis) from 2015 - 2023 in Peking Union Medical College Hospital were enrolled retrospectively. 3345 colonoscopy images were collected. The dataset was split into training (70%) and testing (30%) sets. A multimodal DAFT model integrating clinical biomarkers and endoscopic images was constructed, along with Resnet and Senet models. Model performance was evaluated using accuracy, sensitivity, specificity, positive and negative predictive values from the confusion matrix. Results: UC patients with CMV colitis had distinct clinical characteristics. The multimodal DAFT model outperformed Resnet and Senet in distinguishing UC with CMV colitis, with higher accuracy (0.91), sensitivity (0.87), and specificity (0.93). Conclusion: AI application offers a promising way to enhance early identification of UC with CMV colitis. The multimodal model combining clinical and endoscopic data can assist clinicians in accurate and timely diagnosis.

PubMed Disclaimer

Figures

The image displays three different types of gastrointestinal inflammation: UC+CMV, UC, and normal.
Figure 1.
The endoscopic classification of each category. UC + CMV, UC concomitant with CMV colitis. CMV, cytomegalovirus; UC, ulcerative colitis.
Disease “diabetes”, 20000 “study” published, showing “result” for 24 of 6000 people
Figure 2.
Study flow diagram.
Convolutional feature fusion and classification.
Figure 3.
Deep learning architecture.
Clinical feature importance analysis for diagnosing different pathological conditions.
Figure 4.
Healthy pathological classification results: distinguish healthy samples from pathological samples.
“UC classification results: distinguish UC from healthy and CMV colitis.”
Figure 5.
UC classification results: distinguish UC from healthy and CMV colitis. CMV, cytomegalovirus; UC, ulcerative colitis.
UC+CMV classification results: distinguish CMV colitis from healthy and UC.
Figure 6.
UC + CMV classification results: distinguish CMV colitis from healthy and UC. CMV, cytomegalovirus; UC, ulcerative colitis.
The ROC curve reflects the ability of the model to diagnose classes accurately; micro-average ROC curve shows an area of 0.91, indicating a high diagnostic ability.
Figure 7.
Clinical feature importance by diagnosis class.
Plot of “multi-class ROC” with true positive rate plotted against false positive rate.
Figure 8.
Performance curves. Receiving operating characteristics (ROC) for the testing set.
See this image and copyright information in PMC

Similar articles

See all similar articles

References

    1. Yang H, Zhou R, Bai X, et al. Trend and geographic variation in incidence and prevalence of inflammatory bowel disease in regions across China: a nationwide employee study between 2013 and 2016. Front Med (Lausanne) 2022; 9: 900251. - PMC - PubMed
    1. Beaugerie L, Rahier JF, Kirchgesner J. Predicting, preventing, and managing treatment-related complications in patients with inflammatory bowel diseases. Clin Gastroenterol Hepatol 2020; 18(6): 1324. - PubMed
    1. Oh SJ, Lee CK, Kim YW, et al. True cytomegalovirus colitis is a poor prognostic indicator in patients with ulcerative colitis flares: the 10-year experience of an academic referral inflammatory bowel disease center. Scand J Gastroenterol 2019; 54(8): 976–983. - PubMed
    1. Zhang M, Bai X, Zhang H, et al. The role of cytomegalovirus colitis on short- and long-term outcomes for patients with ulcerative colitis. Scand J Gastroenterol 2022; 57(3): 282–289. - PubMed
    1. Kambham N, Vij R, Cartwright CA, et al. Cytomegalovirus infection in steroid-refractory ulcerative colitis: a case-control study. Am J Surg Pathol 2004; 28(3): 365–373. - PubMed

LinkOut - more resources