- February 01, 2022
Each year around two million women around the world are diagnosed with breast cancer. Typically, tissue samples of the tumor are taken and analyzed and then the cancer is placed into one (1) of three (3) categories – low risk, or grade 1; medium risk, or grade 2; and high risk, or grade 3. The specialists then design a plan of treatment based on this initial assessment.
However, due to its relative lack of detail, this method can lead to patients being treated incorrectly. Now, a team of researchers at Karolinska Institutet in Sweden have developed an AI-based imaging tool that can help specialists to diagnose breast cancer tumors more accurately. “Roughly half of breast cancer patients have a grade 2 tumor, which unfortunately gives no clear guidance on how the patient is to be treated,” said the study’s first author Yinxi Want, Doctoral Student at the Department of Medical Epidemiology and Biostatistics, Karolinska Institutet. “Consequently, some of the patients are over-treated with chemotherapy while others risk being under-treated. It’s this problem that we’ve tried to resolve.”
The team trained the AI to recognize the characteristics of differing classifications of tumor using high-resolution microscopic images taken from 2,800 patients. As a result, it was able to divide patients with grade 2 tumors into two sub-groups – high risk and low risk – enabling specialists to design treatment plans more suited to individual patients. “One big advantage of the method is that it’s cost-effective and fast, since it’s based on microscope images of dyed tissue samples, which is already part of hospital procedure,” said co-author Johan Hartman, Professor of Pathology at the Department of Oncology-Pathology, Karolinska Institutet, and pathologist at the Karolinska University Hospital. “It enables us to offer this type of diagnosis to more people and improves our ability to give the right treatment to any one patient.”
The researchers now plan to further refine the AI and hope to have a fully functioning diagnostic product on the market by 2022.
REFERENCE: Science Focus (the home of BBC science focus magazine); 30 SEP 2021; Jason Goodyer