PRIMUS

Accelerating MS clinical research by several years using synthetic data

Challenges

• Difficult access to data: Years of legal negotiations to access sensitive clinical trial data (Biogen, Merck).

• Open Data Impasse: Impossible to share real data on patients with Multiple Sclerosis (MS) without major risks of re-identification.

• Complexity of the pathology: Need to maintain good statistical fidelity to train a reliable decision support tool (CDSS).

‍

Maintains statistical quality and utility

• Maintaining quality: The synthetic data generated maintains the same statistical properties as the real data.

‍

ROI

• Time-to-data: Passage from several years of administrative blockages to A few weeks To obtain usable data.

• Launch of the CDSS: Creation of a clinical decision support tool that is transforming the management of MS. By exploiting synthetic “placebo arms” from thousands of patients, this tool allows neurologists to choose the most effective treatments sooner, thus effectively improving patients' quality of life through precision medicine.
• Scientific Paper : https://www.jmir.org/2025/1/e71297
‍

“Octopize provides a central answer: using synthetic data makes it possible to remove an obstacle that is otherwise impassable for a clinician. This removes any doubt about confidentiality while respecting the message behind the data.” - Prof. Gilles Edan, Head of the Neurology Department at @CHU in Rennes

“We will provide the neurologist with the experience of 15,000 patients. It's a change of scale, an exceptional revolution.” - Prof. Gilles Edan, Head of the Neurology Department at @CHU in Rennes.

Over 15,000 patient data: The Avatar method has made it possible to gather and anonymize data from more than 15,000 patients from international clinical trials and cohorts, transforming a locked data pool into a strategic asset for medical AI.

Download the use case