Technion
have developed an
artificial intelligence
model aimed at assessing whether early-stage breast cancer patients are likely to benefit from chemotherapy. This decision has historically been challenging for both healthcare professionals and patients.
Chemotherapy following surgery can eliminate residual cancer cells and diminish the likelihood of recurrence. However, for many patients, it may not enhance outcomes and can lead to significant side effects, including long-lasting harm to the immune system. Consequently, oncologists are faced with the challenge of balancing uncertain benefits against potential risks.
3 View gallery
The artificial intelligence system analyzes the digital tissue sample of breast cancer taken at diagnosis
breast cancer
each year, while the figures rise to about 300,000 in the United States and approximately 2.3 million worldwide. Although current tools such as genomic tests like Oncotype DX estimate recurrence risk and the effectiveness of chemotherapy, they tend to be expensive, require weeks for processing, and are not broadly accessible. Additionally, their predictive accuracy is limited.
The innovative AI approach, recently published in The Lancet Oncology and presented at the European Society for Medical Oncology conference, analyzes standard pathology samples taken at diagnosis. Utilizing high-resolution images of tumor tissue, the system assesses patterns correlated with cancer behavior, including cell structure, division rates, and immune response indicators.
“The system examines various regions of the tumor and its surroundings, identifying visual patterns linked to the biological behavior of the cancer,” said Dr. Gil Shamai, who led the study at Technion’s Geometric Image Processing Laboratory.
Instead of depending solely on genetic profiling, the model extracts what researchers refer to as a “visual signature” from stained tissue images. Prof. Ron Kimmel, who heads the laboratory, likens this method to recognizing eye color by observation rather than through genetic analysis.
The outcome is a numerical score that can be generated in a matter of minutes, providing estimates for both the likelihood of recurrence and the potential advantages of chemotherapy. This tool is designed to assist, not to replace, clinical decision-making.
Despite its effectiveness, the system functions as a “black box,” implying that its internal decision-making processes remain unclear. Nevertheless, the team reported consistent predictions across various hospitals and patient groups.
This model was trained on extensive datasets of digital pathology images and subsequently validated using samples from the TAILORx trial, a major randomized clinical study involving over 10,000 breast cancer patients. Such trials are regarded as the gold standard in clinical research due to their ability to reduce bias in treatment comparisons.
Prof. Dvir Aran, a co-leader of the study, emphasized that this validation signifies a milestone: “This is the first artificial intelligence model demonstrated to predict treatment benefits in breast cancer based on pathology samples,” he stated.
Further validation was conducted using patient data from hospitals in Israel—including Carmel, Emek, and Sheba medical centers (which is Israel’s largest hospital, located near Tel Aviv)—as well as institutions in the United States and Australia.
Researchers believe the system could broaden access to personalized medicine, particularly in developing nations where genomic testing is often limited. Since it utilizes existing biopsy samples and standard digital scanning, it does not necessitate extra procedures or specialized laboratory facilities.
The team is working towards clinical implementation in Israel and is planning additional trials in Brazil and India. They are also investigating possible applications for other types of cancer and treatment decisions.
Collaboration with researchers and clinicians from several international institutions, including the Dana-Farber Cancer Institute in Boston and the University of Chicago Medical Center, has been pivotal to the project. Support has also come from various Israeli research initiatives, including the Israel Innovation Authority and the Israel Cancer Research Fund.
Researchers assert that AI-driven tools such as this could become integral in oncology care, facilitating more tailored treatments while minimizing unnecessary interventions.
