Alexandra Enders
Pronouns: she/her
Research Mentor(s): Flora Rajaei
Research Mentor School/College/Department: DCMB / Medicine
Program:
Authors: Alexandra Enders, Flora Rajaei, Flora Rajaei
Session: Session 5: 2:40 pm – 3:30 pm
Poster: 11
Abstract
Crohn’s disease is a chronic bowel disorder that presents challenges in both diagnosis and treatment due to its complex and heterogeneous nature. Our research explores the application of image processing techniques integrated with machine learning algorithms to assist clinicians in diagnosing the severity of the disease more quickly, efficiently, and accurately. In the study, we processed a dataset that was collected from healthy individuals and those who were diagnosed with Crohn’s disease using colonoscopy videos. Utilizing various Python libraries, we processed this dataset to reduce noise, normalize values, and select relevant features. This data was then used as input for our machine-learning models, specifically Logistic Regression and Random Forest algorithms. To evaluate our models, we conducted multiple performance tests, including assessments of AUC (Area Under the Curve), sensitivity, and specificity. These tests confirmed that our models are effective in predicting the severity score of Crohn’s disease.
