Breadcrumb

null Promising development in the fight with cancer using artificial intelligence

New privacy-preserving approach uses hospital based medical notes to refine predictions and prognosis in cancer patients over time

Source: University of California – San Francisco and RI-MUHC. A new study published on July 22 in Nature Cancer demonstrates how an artificial intelligence framework can integrate longitudinal electronic health records with real-world data and help to improve and personalize cancer treatments. With collaborators in other Canadian, American and European research centres, Jan Seuntjens, PhD, of the Research Institute of the McGill University Health Centre (RI-MUHC), is part of the international consortium developing this novel approach, along with his former postdoctoral fellows Martin Vallières, PhD, now at the Université de Sherbrooke, and Avishek Chatterjee, PhD, now at Maastricht University.

Jan Seuntjens, PhD, is part of the MEDomics consortium and a member of the Cancer Research Program at the Research Institute of the MUHC
Jan Seuntjens, PhD, is part of the MEDomics consortium and a member of the Cancer Research Program at the Research Institute of the MUHC

"We reported results of the first longitudinal approach to natural language processing of unstructured medical notes and demonstrated its ability to update and improve a prognostic model over time, as a patient's oncologic illness course unfolds,” says Olivier Morin, PhD, leader of the project and Chief of Physics in the Department of Radiation Oncology at the University of California San Francisco (UCSF).

Catherine Park, MD, co-senior author and chair of the Department of Radiation Oncology at UCSF, added: "With this data we were able to validate findings of published clinical trials using real world data: for example, the positive impact of immunotherapy in lung cancer. In addition, there are exciting opportunities to generate hypotheses based on associations from patients' individual health profiles, and risk factors."

The consortium has created a secure, dynamic, continuously learning and expandable infrastructure, termed MEDomics, designed to constantly capture multimodal electronic health information, including imaging, across a large and multicentric healthcare system. The team has created this animation to explain the concept.

Professor Phillippe Lambin, senior author and chair of the Department of Precision Medicine at Maastricht University, adds, "The MEDomics infrastructure allowed us to validate several new clinical hypotheses like the importance of cardiovascular morbidity for the outcome of cancer treatment."

Part of the code is open source, and the team of collaborators would like to expand the international consortium, encouraging interested parties to visit www.medomics.ai.

“Our vision is to create an open-source computation platform integrating all MEDomics developments and from which both clinical staff and research scientists could tackle a diverse range of oncological problems using AI,” says Olivier Morin.

Adds Jan Seuntjens, “Electronic multimodal health data currently lives in multiple, segregated systems across our healthcare system. This study shows the enormous potential when real-world multimodal health data can be integrated, thereby enabling continuous learning while simultaneously capturing and integrating longitudinal clinical records. We hope to leverage this technology to enable a wide variety of studies in Canadian healthcare centres.”

Read the publication in Nature Cancer and a commentary by researchers at Memorial Sloan Kettering Cancer Center in the same issue.

Media Contact:
Olivier Morin, PhD
Olivier.Morin@ucsf.edu

August 5, 2021