Radiomic Prognostic Vector IDs patients with median overall survival of less than two years
FRIDAY, Feb. 15, 2019 (HealthDay News) — Artificial intelligence software can predict prognosis in epithelial ovarian cancer (EOC) above established prognostic methods, according to a study published online Feb. 15 in Nature Communications.
Haonan Lu, from the Ovarian Cancer Action Research Centre at Imperial College London, and colleagues extracted 657 quantitative mathematical descriptors from the preoperative computed tomography images of 364 EOC patients at their initial presentation. A noninvasive summary statistic of the primary ovarian tumor was derived using machine learning, based on four descriptors, and named Radiomic Prognostic Vector (RPV).
The researchers found that RPV was able to reliably identify the 5 percent of patients with median overall survival of less than two years, significantly improving established prognostic methods. RPV was validated in two independent multicenter cohorts. Stromal phenotype and DNA damage response pathways were activated in RPV-stratified tumors as elucidated by genetic, transcriptomic, and proteomic analysis from two independent datasets.
“The long-term survival rates for patients with advanced ovarian cancer are poor despite the advancements made in cancer treatments. There is an urgent need to find new ways to treat the disease,” a coauthor said in a statement. “Our technology is able to give clinicians more detailed and accurate information on the how patients are likely to respond to different treatments, which could enable them to make better and more targeted treatment decisions.”
One study author is an employee of AstraZeneca.
Copyright © 2019 HealthDay. All rights reserved.
