Improving ovarian cancer care through AI
A team of research fellows led by Professor Geoff Hall is developing AI tools to support ovarian cancer research in Leeds. The work aims to deliver the best possible care for ovarian cancer patients and is made possible thanks to a generous donation from Yorkshire businessman, Tony Bramall and his family, in memory of his daughter Melanie Foody who passed away in 2019 after a year-long battle with the disease.
A Medical Oncologist and Professor of Digital Health and Cancer Medicine at the Leeds Institute of Medical Research, Prof Hall is also a senior clinical leader at the Leeds Institute of Data Analytics (LIDA). His interest in real-world data, and his work with the School of Computing and Centre for Doctoral Training in AI in Healthcare, meant he was uniquely positioned to develop a programme for multi-modal AI in ovarian cancer that would link data from electronic health records, including digital pathology, imaging, genetic and molecular profiling.
“The Leeds Cancer Centre has collected detailed data for more than 20 years on patients with ovarian cancer, including diagnosis, treatment and outcomes,” says Prof. Hall. “I have a team enhancing that data to ensure it is capable of supporting collaborative research with other centres across the world.”
The data was recognised as capable of supporting a number of important research programmes – and, with the gift from the Tony Bramall Charitable Trust we have been able to appoint clinical research fellows in pathology and radiology, a senior research fellow working with genetic data from Genome England and to fund expert support from within the School of Computing. Although each research project can work independently of the others, the overall ambition will bring these different types of data together to generate the deepest insight into better ways to care for ovarian cancer patients.
Unlike a lot of other malignancies, where the decision to undertake surgery and/or the type of surgical procedure is determined after a pathologist looks at the tissue samples, in ovarian cancer, the decision is often made using blood tests and medical imaging findings alone. Imaging findings are not always definitive, and this leads to uncertainty. “By collaborating with the other research fellows, we aim to combine data from imaging, pathology and clinical data in order to further refine our predictions, reducing clinical uncertainty, and get the correct treatment for every patient, earlier detection of disease relapse and chemotherapy resistance.” Adds Dr Pratik Adusumilli, Clinical Research Fellow on the project.
The result of this would allow for better treatment choices, earlier enrolment in clinical trials with alternative therapies and, in turn, may improve patient outcomes.
The investment has already had significant impact in this new field, with cross-fertilisation occurring between departments and resulting in partnerships being established. The team are also working with three students and their supervisors within the UKRI Centre for Doctoral Training in Artificial Intelligence for Medical Diagnosis and Care.
Prof. Hall stresses that the potential impact of this project is huge. “The AI tools being developed can enable the best possible care for ovarian cancer patients anywhere in the world,” he says. “In the future, we hope to enable clinicians across the world to upload data for analysis by algorithms developed in Leeds to provide guidance on and provide support for decisions made about their care.”
About Professor Geoff Hall
Professor Hall is an honorary consultant at the Leeds Cancer Centre where he specialises in the management with women with gynaecological cancer. His research interests focus on the use of data captured during routine care to improve pathways of care and outcomes for cancer patients of the future. He is also one of the Chief Clinical Information Officers at Leeds Teaching Hospitals, the Chief Clinical Data Officer for Health Data Research UK and a co-director of the University’s Centre for Doctoral Training in Artificial Intelligence for Medical Diagnosis and Care.