Developing a Multi-parameter Prognostic Prediction Model for Disability Ranks and Progression of Patients with Multiple Sclerosis at the Early Stages of the Disease

Integrating clinical, imaging, and biological data from large real-world MS cohorts, this project builds a machine learning model to predict disability progression early and enable personalized treatment decisions.
Ongoing Multiple Sclerosis Neurologic Diseases Data-Driven & Quantitative Disease Subtyping Early Disease Detection Experimental Platforms & Models Human Cohorts Machine Learning & AI Precision Medicine Real-world Evidence Translational & Clinical

Funding Year: 2023

Project Overview

Multiple sclerosis is a chronic neurological disease that can lead to significant long-term disability, yet current tools offer limited ability to predict which patients will progress most severely. This project develops a predictive model trained on large, real-world datasets from MS patients, integrating clinical, demographic, imaging, and biological data. Using advanced statistical techniques and machine learning, the team aims to generate accurate, personalized prognoses early in the disease course — enabling clinicians to match therapy intensity to individual risk, prevent unnecessary side effects, and reduce long-term disability. With robust validation, the tool is designed for clinical implementation within 3–5 years.

Impact & Innovation

Predicting MS disability before it advances.

 

By integrating clinical, imaging, biological, and demographic data from thousands of MS patients, this project builds a machine learning tool that personalizes prognosis early — when treatment decisions matter most.

  • Reveals novel biological and clinical patterns driving MS progression, deepening mechanistic understanding of the disease
  • Generates IP potential through a patentable digital health tool for individualized risk stratification and treatment guidance
  • Advances the Consortium’s Integrated Data and Discovery Platforms pillar by demonstrating how multi-modal real-world data can drive precision prognostics in neurological autoimmune disease
Research Approach

A framework designed for discovery

This project applies advanced statistical modeling and machine learning to large, real-world multi-modal datasets to develop and validate a comprehensive prognostic tool for early disability prediction in multiple sclerosis. The work integrates diverse data types to build a model capable of supporting individualized clinical decision-making.

Advanced statistical techniques and machine learning algorithms applied to high-dimensional, real-world MS patient data, integrating clinical, demographic, imaging, and biological variables to construct and validate a multi-parameter prognostic prediction model.

Large real-world datasets from MS patients encompassing clinical assessments, demographic profiles, neuroimaging data, and biological markers, analyzed longitudinally to identify predictors of disability progression and inform model development.

Development of a validated, personalized prognostic model that enables early identification of MS patients at high risk of disability progression, with a translational goal of clinical implementation within 3–5 years to support precision treatment decisions and reduce long-term disease burden.

Investigators & Institutions

Powering the science

Principal Investigator

Arnon Karni, MD, Colton Consortium Member

Associate Professor, School of Continuing Medical Education (Neurology), Tel Aviv Sourasky Medical Center, Tel Aviv University

Key Collaborators

Malka Gorfine, PhD, Colton Consortium Member

Professor, Statistics and Operations Research, Tel Aviv University

Hadar Kolb, MD, Colton Consortium Member

Head of Precision Medicine, Neuroimmunology, Neurology, Tel Aviv Sourasky Medical Center, Tel Aviv University

Colton Centers & Institutions

Colton Center for Autoimmunity at Tel Aviv University