Revolutionary Machine Learning Classifier Accurately Identifies Ovarian Cancer Through Metabolic Profiles

Innovative tool utilizes serum metabolic profiles to distinguish between cancerous and non-cancerous individuals

June 11, 2024  –  The Ovarian Cancer Institute (OCI) announces the development of a groundbreaking machine learning-based classifier capable of identifying ovarian cancer with remarkable accuracy. This innovative tool utilizes serum metabolic profiles to distinguish between cancerous and non-cancerous individuals, offering a new direction in early cancer detection.

A study was to develop a machine learning-based classifier that leverages metabolic profiles of serum samples to accurately identify individuals with ovarian cancer. Researchers analyzed serum samples from 431 ovarian cancer patients and 133 healthy women across four geographic locations using mass spectrometry. Through recursive feature elimination and repeated cross-validation, reliable metabolites were identified and used to create a consensus classifier. This classifier assigns probabilities to individual samples, indicating the likelihood of cancer presence.

The consensus classification model distinguished cancerous from non-cancerous samples with an impressive 93% accuracy. The individual patient scores, derived from the model, facilitated the development of a clinical tool that assigns a likelihood of cancer presence or absence.

Benedict B. Benigno, MD, the Founder and CEO of OCI, highlighted  this integrative approach, which combines metabolomic profiles with machine learning-based classifiers, has led to the development of a clinical tool offering a probabilistic assessment of ovarian cancer. This method surpasses traditional binary tests in both clinical informativeness and accuracy, representing a significant advancement in early ovarian cancer detection.

Key Study Highlights:

• Predictive models derived from machine learning analyses of serum metabolic profiles can accurately detect ovarian cancer.
• Only a minority of the most predictively informative metabolites is currently annotated (7%).
• Lipids predominate among the most predictively informative metabolites currently annotated.
• The frequency distribution of model-derived patient scores were used to develop a clinical tool for the diagnosis of OC.

Jeffery Skolnick, PhD, Chief Scientific Officer, stated, “This personalized and probabilistic approach to cancer diagnostics is a promising advancement, offering a more accurate and clinically informative alternative to traditional methods. Early detection is crucial in the fight against ovarian cancer, and this tool represents a significant step forward.”

About the Ovarian Cancer Institute:

The Ovarian Cancer Institute is dedicated to advancing research, education, and early detection strategies for ovarian cancer. Through cutting-edge research and innovative approaches, we strive to improve the lives of those affected by ovarian cancer. Learn more: https://www.ovariancancerinstitute.org

Review the Study:

https://www.sciencedirect.com/science/article/pii/S0090825823016360

###

Contact:

Kristin Krancer, MPH

Executive Director
Ovarian Cancer Institute
kkrancer@ovariancancerinstitute.org
404-666-0317