The image, provided by Yonhap News TV, shows a CT scanner room.
SEOUL, July 12 (Korea Bizwire) — A new artificial intelligence (AI) algorithm has been developed that can predict lung cancer patients’ response to immuno-oncology treatment on the basis of their clinical information.
A joint research team from the Yonsei Cancer Hospital and the Yonsei University Medical School said it had developed a machine learning-based algorithm that can predict a non-small cell lung cancer (NSCLC) patient’s response to immune-oncology treatment.
NSCLC patients, who account for more than 80 percent of all lung cancer patients, are typically tested for programmed death-ligand 1 (PD-L1), a protein on the surface of the cancer cell, prior to undergoing immuno-oncology treatment.
This is aimed at predicting their response to such treatment.
Currently, the PD-L1 test has an accuracy rate of 64 percent in predicting patients’ response to such treatment.
The research team developed an algorithm to predict patients’ response to immuno-oncology treatment using data from 142 NSCLC patients who underwent treatment using the PD-L1 index at Severance Hospital.
The algorithm reflects 19 different kinds of clinical data, including the existing PD-L1 expression level, age, gender, tumor size, transferred location, and general blood test results.
The algorithm showed an accuracy rate of 82 percent, far higher than those of existing diagnosis kits designed to predict patients’ response to the immuno-oncology treatment.
J. S. Shin (js_shin@koreabizwire.com)
