Classification and prediction of inflammatory bowel diseases

Abstract

The project aims to develop a system For predicting inflammatory bowel and building the machine learning model based on routinely performed laboratory blood, urine, and fecal tests to supportdifferentiation between IBD patients and non- IBD patients comparison of the effectiveness of our model to standard inflammatory serummarker, that is C-reactive protein (CRP), in the prediction of IBD, creating a website- based application supporting the prediction of the presence of IBD . the age profile of IBD patients is changing and there is an increase in early-onset and late-onset IBD prevalence. Both groups (older adults, which frequently suffer from various comorbidities, as well as children) would particularly benefit from the non-invasive diagnostic test. However, colonoscopy remains a gold standard in IBD diagnosis, monitoring of the disease course, and response to the therapy, as well as colorectal cancer screening [6,7,8,9]. Still, despite its obvious advantages, it is highly invasive, expensive, time- consuming, requires qualified medical personnel and patient‘s preparation, and is often poorly tolerated by patients themselves. Besides, in the pandemic all low-contact medical procedures are preferred. creating a website- based application supporting the prediction ofthe presence of IBD.Therefore, a simple diagnostic methodology based only on markers from blood, urine and stool that can be performed by a GP would be imperative inthe early diagnosis of IBD Keywords : inflammatoryboweldisease;ulcerative colitis; Crohn’s disease; artificial intelligence; machine