Cancer prediction tool combines machine learning, radiomics

Researchers at Mount Sinai and USC have developed a predictive framework that can distinguish between low- and high-risk prostate cancer.

The prediction tool combines machine learning with radiomics, a branch of medicine that uses algorithms to extract large amounts of quantitative characteristics from medical images. Researchers say it can tell different risks apart with unprecedented accuracy.

“By rigorously and systematically combining machine learning with radiomics, our goal is to provide radiologists and clinical personnel with a sound prediction tool that can eventually translate to more effective and personalized patient care,” says Gaurav Pandey, assistant professor of genetics and genomic sciences at the Icahn School of Medicine at Mount Sinai.

Also See: Machine learning detects lymphedema in breast cancer survivors

Multi-parametric magnetic resonance imaging—which detects prostate lesions—and a five-point scoring system that classifies lesions found on the mpMRI are currently used for the clinical assessment of prostate cancer.

However, the method is subjective in nature and does not distinguish clearly between intermediate and malignant cancer levels, leading to differing interpretations among clinicians.

“The pathway to predicting prostate cancer progression with high accuracy is ever improving, and we believe our objective framework is a much-needed advancement,” adds Pandey.

In a paper published last week in Scientific Reports, Pandey and his colleagues present results of a single institution, retrospective study involving prostate cancer patients at the USC Keck Medical Center.

“Radiomics, in combination with machine learning, can help achieve objective classification of clinical images that can be a valuable tool to aid clinicians in identifying appropriate treatment options for patients without subjecting them to unnecessary intervention,” state the authors.

“The goal of incorporating machine learning into radiomics is not to compete with the radiologist, but to rather provide the radiologist and physician team taking care of the patient with objective prediction tools that can aid personalized decision making regarding individual disease course and treatment outcome,” they add.