February 4, 2023
February 3, 2023

AI Able to Predict when Lesions in the Breast will Develop into Invasive Cancer

Reviewed By:
AI Able to Predict when Lesions in the Breast will Develop into Invasive Cancer

A team of scientists at Case Western Reserve University have put together an Artificial Intelligence (AI) that could predict whether precancerous breast lumps will later evolve into invasive cancer. If proven effective, this tool could prevent unnecessary rounds of radiation for would-be breast cancer patients. According to experts from Case Western, women with precancerous lumps in the breast will often elect to receive surgery to get rid of any remaining tissue and often even undergo radiation therapy, even if they’re at a low risk of developing metastasized cancer.

Professor of biomedical engineering at Case Western, Anand Madabhushi, told the university’s news service that, “in short, we’re probably overtreating patients. That goes against prevailing wisdom, but that’s what our analysis is finding.”

That’s where this team, led by Jaojia Li came in: they analyzed lumpectomy tissue from around 60 patients who were diagnosed with “Stage 0 breast cancer,” which is also called ductal carcinoma in situ (DCIS). About 20% of breast cancers are diagnosed at this stage; that accounts for about 60,000 Americans. Li observed that tumors that were generally labeled as “intermediate” were actually closer to being at low-risk of recurrence in not only orientation but also size. She distinguished and validated between high- and low-risk groups of tumors, then was able to predict the likelihood of a tumor to progress from Stage 0 to invasive cancer.