MEDICAL NEWS: Preventing the harm of immunotherapy through artificial intelligence

【British Medical Express website February 5, 2021 article】 Title: Prevent the harm caused by immunotherapy through artificial intelligence.

Researchers at Case Western Reserve University in the United States used artificial intelligence (AI) to analyze simple tissue scans and claimed to have found a biomarker that could tell doctors which lung cancer patients' condition might be aggravated by immunotherapy.

Researchers and oncologists divide lung cancer patients into two broad categories: those who will benefit from immunotherapy and those who may not benefit from immunotherapy, while a third category of lung cancer patients has begun to emerge, known as "over-progressioners," a category that is actually harmed by immunotherapy (including shortened lifespan after treatment).

The study shows that artificial intelligence and machine learning can predict which lung cancer patients will benefit from immunotherapy. In this and previous research, scientists have worked to teach computers to look for and identify patterns of CT scans when lung cancer is first diagnosed to reveal useful information known before treatment.

Madabhushi, senior author of the study, said the study's importance lies in demonstrating that the radiological patterns of conventional CT scans can identify the responses of lung cancer patients receiving immunotherapy: responders, non-responders, and highly progressives. In addition, the scientists found the most important clues: the characteristics of radiological substances outside the tumor are more predictive than the characteristics of the radiological substances inside the tumor, and the changes in the blood vessels around the nodules are more predictive, so that it is possible to know which patients will be harmed by extra-tumor immunotherapy.

Original article, please indicate the source when reprinting