Deciding whether to administer chemotherapy after surgery is one of the most challenging questions in early‑stage breast cancer care. While chemotherapy can reduce the risk of recurrence, most patients do not benefit from it and may experience significant short‑ and long‑term side effects. The central challenge is identifying, at the time of diagnosis, which patients are likely to benefit and which are not.
Researchers from the Technion - Israel Institute of Technology, together with collaborators from leading medical centers in the United States and Europe, have developed an artificial intelligence (AI) model that predicts both the risk of breast cancer recurrence and the likelihood that a patient will benefit from chemotherapy. The model analyzes routine pathology slides taken at diagnosis, offering a fast, widely accessible alternative to costly genomic tests.
The study was recently published in The Lancet Oncology and presented at the European Society for Medical Oncology (ESMO) conference. It is the first AI model of its kind to be validated in a large, randomized clinical trial.
Addressing a Global Clinical Need
Each year, approximately 2.3 million people worldwide are diagnosed with breast cancer, including about 300,000 in the United States and 5,000 in Israel. Today, genomic tests such as Oncotype DX are commonly used to guide chemotherapy decisions, but these tests are expensive, can take weeks to return results, and are unavailable to many patients globally. Their predictive accuracy is also limited, leading to both unnecessary chemotherapy and missed treatment opportunities.
The Technion-led AI model aims to address these limitations by using information already available in standard pathology samples.
How the Model Works
The system analyzes high‑resolution digital images of tumor tissue stained and examined as part of routine pathology. Using deep learning, it evaluates multiple regions of the tumor and its microenvironment, identifying visual patterns associated with cancer behavior, including cell division, tissue structure, immune response, and features linked to treatment sensitivity or resistance.
“These are complex biological signals that the human eye cannot consistently quantify,” said Dr. Gil Shamai of the Technion’s Geometric Image Processing Laboratory, who led the study. “The model integrates many subtle cues to generate a score that reflects both recurrence risk and expected benefit from chemotherapy.”
Prof. Ron Kimmel, head of the laboratory in the Henry and Marilyn Taub Faculty of Computer Science, explained the concept: “Instead of testing genes, we look directly at the tissue. Just as eye color can be determined by looking at the eyes rather than analyzing DNA, our system extracts a visual signature from pathology images that informs optimal treatment.”
Clinical Use and Validation
Clinically, the process is straightforward. After diagnosis, the existing tissue sample is digitally scanned and securely analyzed by the AI system. Within minutes, the model produces a numerical score that supports shared decision‑making between oncologist and patient.
While the system’s internal decision‑making cannot be fully explained in simple rules, its performance has been rigorously validated. The researchers were granted rare access to tissue samples and clinical data from the TAILORx trial – one of the largest randomized breast cancer studies, involving more than 10,000 patients who were randomly assigned to receive chemotherapy or not.
“Using data from a randomized trial allowed us to test whether the model truly predicts benefit from chemotherapy, not just recurrence risk,” said Dr. Shamai.
According to Prof. Dvir Aran of the Technion’s Faculty of Biology, a co‑leader of the study, “This is the first AI model shown to predict treatment benefit in breast cancer directly from pathology samples.”
The model was further validated on thousands of patients from hospitals in Israel, the United States, and Australia, including Carmel, Emek, and Sheba Medical Centers, demonstrating consistent performance across different populations, equipment, and healthcare systems.
Fast, Affordable, and Globally Scalable
Unlike genomic tests, the AI‑based assessment requires no additional tissue, laboratory processing, or waiting period. It can be performed in minutes in any pathology lab equipped with a digital scanner and internet access.
“In developing countries, where genomic testing is largely unavailable, this tool could dramatically expand access to personalized cancer care,” said Prof. Aran. “In high‑income countries, it could reduce costs, shorten diagnosis time, and improve decision accuracy.”
Looking Ahead
The research team is now advancing steps toward clinical implementation in Israel and preparing clinical trials in Brazil and India, where the potential impact is particularly large. The researchers are also working to further improve the model and extend it to additional treatments and cancer types where aggressive therapy decisions are made under uncertainty.
Based on these impressive results and the knowledge accumulated over years of groundbreaking research, the researchers now intend to establish a company that will develop tests making them significantly more accessible, accurate, and faster compared to those currently in use worldwide.
The study was led by Dr. Gil Shamai, Prof. Ron Kimmel, and Prof. Dvir Aran, in collaboration with oncologists and pathologists from institutions including Dana‑Farber Cancer Institute, Mount Sinai Medical Center, the University of Chicago Medical Center, and IPATIMUP Medical Center in Portugal.
The research was supported by the Israel Innovation Authority, the Zimin Institute for AI Solutions in Healthcare at the Technion, the Israel Personalized Medicine Partnership, and the Israel Cancer Research Fund.
The Lancet Oncology
Randomized controlled/clinical trial
People
Deep learning on histopathological images to predict breast cancer recurrence risk and chemotherapy benefit: a multicentre, model development and validation study
11-Mar-2026