Development and Validation of a Fair and Interpretable Machine : Learning Model for Non-Invasive Diagnosis of At-Risk MASH Using Routine Clinical and Biochemical Data / Christina Al Nabbout

0 review

Development and Validation of a Fair and Interpretable Machine : Learning Model for Non-Invasive Diagnosis of At-Risk MASH Using Routine Clinical and Biochemical Data / Christina Al Nabbout

Study and report

Al Nabbout, Christina | Rezaei, Mostofa. Encadrant académique

Published by ESCP Business School - 2025

Library: Berlin London Paris Champerret Paris Montparnasse Turin Madrid

Source: Syracuse

Item type

Study and report

Dissertation

MBD Big Data & Business Analytics

Language

English

Publication date: 2025

Subjects: ARTIFICIAL INTELLIGENCE - MEDICINE

Faced with the global rise of metabolic liver diseases, one key question emerges: can we diagnose effectively without relying on invasive procedures like liver biopsy? This thesis explores a bold yet practical answer: using artificial intelligence to fully leverage already available medical data. Based on simple clinical and biochemical parameters, a machine learning model is developed to identify patients at risk of MASH, the most severe form of the disease. Blending scientific rigor with clinical applicability, this work proposes an accessible, transparent, and equitable alternative to traditional diagnostic tools. Rather than replacing advanced technologies, it aims to democratize screening by making better use of what healthcare systems already have. A modern reflection on the future of medicine—where innovation lies not only in complexity, but in the ability to make more out of the existing.