Serological Biomarker-Based Machine Learning Models for Predicting the Relapse of Ulcerative Colitis

• DOI: 10.2147/jir.s423086
• Publication Date: 2023-08-21T08:05:29Z
• Authors: Wenwen Pang; Bowei Zhang; Leixin Jin; Yao Yao; Qiurong Han; Xiaoli Zheng
• Keywords: relapse; serological markers; machine learning; random forest model; Pathology; RB1-214; Therapeutics. Pharmacology; RM1-950; Journal of Inflammation Research; 3. Good health; ulcerative colitis; Original Research

Purpose: To explore whether machine learning models using serological markers can predict the relapse of Ulcerative colitis (UC). Patients and Methods: This clinical cohort study included 292 UC patients, were obtained when patients discharged from hospital. Subsequently, four including random forest (RF) model, logistic regression decision tree, neural network compared to UC. A nomogram was constructed, performance these evaluated by accuracy, sensitivity, specificity, area under receiver operating characteristic curve (AUC). Results: Based on patients' characteristics markers, we selected relevant variables associated with developed a LR model. The novel model gender, white blood cell count, percentage leukomonocyte, monocyte, absolute value neutrophilic granulocyte, erythrocyte sedimentation rate established for predicting relapse. In addition, average AUC 0.828, which RF best. test group 0.889, accuracy 76.4%, sensitivity 78.5%, specificity 76.4%. There 45 in models, relative weight coefficients determined. Age has greatest impact classification results, followed hemoglobin concentration, platelet distribution width. Conclusion: Machine based had high be used noninvasively patient outcomes an effective tool determining personalized treatment plans. Keywords: ulcerative colitis, relapse, learning,