Cancer Research Horizons and Royal Surrey NHS Foundation Trust have announced a collaboration with Health Triage, an Italian start-up specialising in AI for medical diagnostics, to advance breast cancer screening using data from the OPTIMAM Medical Image Database.
As part of the collaboration, Health Triage will use OPTIMAM data to train deep learning models to automatically identify mammograms that do not show signs of cancer. Triaging mammograms with a high likelihood of being negative can reduce the workload of radiologists by enabling them to initially focus on suspicious and high-risk cases.
Founded in 2020 by Raffaele Petrone and Davide Dettori, Health Triage is based in Turin - in which the start-up successfully completed an incubation path in I3P, the Innovative Companies Incubator of Politecnico di Torino - and Pozzuoli (Naples). In May 2025, the young company announced the successful closing of a €5 million investment from the ENEA Tech and Biomedical Foundation.
Artificial intelligence in the fight against cancer
Given that around 99.1% of screening mammograms in the UK return a negative result and the shortage of radiologists in many countries, the use of AI-based solutions has the potential to reduce healthcare resource use and optimise radiology workflows, while maintaining breast cancer detection rates.
The use of OPTIMAM data will also enable Health Triage to investigate whether models trained specifically as rule-out systems can achieve better specificity than a system trained for the purpose of cancer detection, which could help further improve screening programmes.
OPTIMAM is a centralised, continuously expanding database of approximately 7 million anonymised mammography images from over 465,000 clients collected from multiple National Health System (NHS) breast screening sites across the United Kingdom. Royal Surrey manages the OPTIMAM database and ongoing image collection with continued funding from Cancer Research UK. OPTIMAM data is available for academic and commercial research, subject to a data sharing agreement with Cancer Research Horizons and Royal Surrey. For further information on the access process, it is possible to visit the OPTIMAM website.
To ensure the clinical validation of Health Triage’s solution, the Italian Group for Mammography Screening (GISMa) will act as an independent body responsible for the scientific oversight of the project.
Technology and innovation for public health
Tony Hickson, Chief Business Officer of Cancer Research Horizons, said: “Breast screening programmes are critical for the early detection of cancer and Health Triage’s use of OPTIMAM data has the potential to improve these while reducing the burden on healthcare systems. The AI revolution in healthcare relies on high-quality data, and this partnership would not be possible without Royal Surrey’s development and management of OPTIMAM, one of the largest and best-curated imaging databases available.”
Davide Dettori, CEO of Health Triage, commented: “Artificial Intelligence is transforming the future of medical diagnostics. Our goal is to make breast cancer screening more accurate, widespread and accessible. Thanks to our collaboration with Royal Surrey NHS Foundation Trust and Cancer Research Horizons, the use OPTIMAM data, and the scientific support of GISMa, we are able to provide tangible support to radiologists and screening programmes – ultimately helping to save more lives.”
Louise Stead, Group Chief Executive of Royal Surrey and Ashford and St Peter’s Hospitals NHS Foundation Trusts, said: “We are proud to support this important collaboration, which has the potential to make a real difference in how breast cancer is detected and diagnosed. The OPTIMAM database is one of the world’s most comprehensive collections of mammography images, and it’s exciting to see it powering innovation to help adopt AI into healthcare. This technology is intended to help patients to get the answers and care they need sooner and ultimately, save lives.”
This partnership aligns with Cancer Research Horizons’ Commercial Data Partnership Guiding Principles, ensuring safe and transparent use of patient-derived data with the primary aim of positive patient impact.
