This exciting project examines whether AI-based technology can help in patient care by helping consultant pathologists to see if breast cancer has spread to the lymph nodes. This has the potential to improve the accuracy and the speed of assessment, with benefits both to the NHS and to patients.Professor Sarah Pinder, from the School of Cancer & Pharmaceutical Sciences
01 February 2022
AI technology investigated for breast cancer diagnosis
A new study led by academics at King’s and Guy’s and St Thomas’ NHS Foundation Trust is using historical scans to evaluate how AI-driven diagnostic software can aid breast cancer assessments.
Breast cancer is the most common cancer worldwide, with around 56,000 new cases every year. When a breast cancer diagnosis is made, a clinician will check whether the cancer has spread beyond the breast tissue. For this, lymph node tissue is examined by expert pathologists to determine if the cancer has spread. The pathologist’s diagnosis determines the treatment program necessary for each patient.
New technology has now been developed by Paige, a digital diagnostics software company. Paige Lymph Node is an AI software system that can assist pathologists in identifying whether cancer in the breast has potentially spread to the lymphatic system.
The study will assess the impact of the software-assisted diagnostic techniques using historical lymph node cases from Guy’s Cancer Centre. This will be on real lymph node images from breast cancer patients, who have given consent for the anonymous images to be used in research.
This study will show how the technology could help pathologists in their decision-making, in which case further clinical studies would investigate the use of this technology in standard care. The technology is currently being used for research only and is not available for use in clinical care.
The research is being co-led by Danny Ruta, AI Clinical Lead at Guy’s and St Thomas’, and Dr Anita Grigoriadis, Reader in the School of Cancer and Pharmaceutical Sciences at King’s College London.
Dr Anita Grigoriadis said: “Developing, evaluating and implementing novel AI-based approaches is a promising area for research. These technologies have the potential to change the future of healthcare. Working with Paige is opening up new ways for us to participate in taking AI-based approaches from design to application.”
Paige Lymph Node has been developed using advanced computational and AI techniques based on a type of AI called deep learning that imitates the way humans gain knowledge. The results are presented to the pathologists to help them identify and classify potential metastatic tissue.
The technology is not designed to replace pathologist assessment, rather the AI is to be used to allow pathologists to work much more efficiently and could help draw their attention to cancers they might have missed.
The system, similar to the FDA market authorised Paige Prostate, the first-ever authorization of an AI-based system in pathology, is designed to detect suspicious tissue robustly and very quickly with results delivered through a clinical-grade viewer.
Professor Sebastien Ourselin, Head of School of Biomedical Engineering and Imaging Sciences at King’s, said: “The ability of AI to provide us with a faster diagnosis of whether a breast cancer has spread to the lymph nodes is very promising. Paige Lymph Node may allow clinicians to more fully understand the disease in each individual and better plan for the treatment pathway. It is only through NHS, industry and university partnerships such as these that we can get this type of treatment into patient care.”
Dr David Klimstra, CMO at Paige: “Pathologists are central to the diagnosis of breast cancer, and their diagnoses strongly impact each patient’s treatment program. The pathology leaders at King’s College London and Guy’s and St Thomas’ are advancing the methods and science to help patients receive their most accurate diagnoses and to make life-changing decisions on treatments in the most efficient way possible. Paige Lymph Node with this study will explore the potential for artificial intelligence to assist pathologists conducting the important diagnostic assessment of lymph node metastases, defining the risk of progression of breast cancer.”
The study is financially sponsored by Paige.AI, LTD.