How can we … develop machine learning tools to map ancient social networks?
19 December 2024
Integrating data to fight cancer: Sharing ideas and best practice
Eleanor Wolmark, Programme Manager of the Mark Foundation Institute for Integrated Cancer Medicine at the CRUK Cambridge Centre
05 December 2024
About half of us will be diagnosed with some form of cancer in our lifetime . Currently, patient data is held in different silos and between different institutions, making it difficult for researchers and clinicians to learn more about cancers and treat them.
The Mark Foundation Institute for Integrated Cancer Medicine (MFICM) at the CRUK Cambridge Centre aims to integrate data that can be used to research different disease types. We focus on cancers that are especially difficult to treat - breast, ovarian, urological malignancies - integrating patient data from across different institutions and databases into a single platform in a collaboration between the University, Cambridge University Hospitals (CUH) and GE Healthcare.
Harnessing the power of AI helps us to integrate all kinds of data and enables clinicians to see the longitudinal journey of the patient they are treating, accessing scans, images, digital pathology and other clinical information in one place, aiming to help them more easily conceive personalised treatment plans. AI is also used within the data stream pipelines – like using algorithms to auto segment images to show all areas of disease for example, thereby saving clinicians time and again helping to better inform treatment decisions.
A symposium to share knowledge
The MFICM hosted a symposium on the use of AI in cancer care to highlight the work that researchers across the University of Cambridge are currently engaged in, and to bring together global experts and cancer patients. Co-hosted by Dr Mireia Crispin-Ortuzar and myself, it took place in July 2024 and was funded by the Accelerate Programme for Scientific Discovery and Cambridge Centre for Data-Driven Discovery.
More than 100 delegates came together at Downing College, and another 400 joined the two-day event online. Delegates comprised global leaders in their areas both international and from the University of Cambridge as well as established and early-career researchers, students, clinicians, surgeons, industry partners, patients and the public. They talked about the opportunities and challenges working with AI in a healthcare setting can bring, with 14 speakers including seven international speakers sharing their areas of expertise.
For example, Faisal Mahmood, Associate Professor, Pathology, Harvard Medical School, Associate Professor, Division of Computational Pathology, Brigham and Women’s Hospital, talked about multimodal and generative AI for pathology. Discussions were centred around AI in precision oncology, including early prediction/detection, integration of multiple data streams to aid clinical decision making and collaboration between industry, academia and the NHS. The real strength of the event was bringing together experts on panels to talk about the benefits of using AI in their area of work. For example, Professor Raj Jena, Clinical Consultant, Department of Oncology at the University of Cambridge talked about the lessons he has learned from deploying a cloud-based open-source imaging AI at Cambridge University Hospitals, while Javier Alvarez-Valle, Senior Director of Biomedical Imaging, Microsoft Research, Cambridge shared information about the MAIRA-2 model and grounded radiology report generation. Bringing these two speakers together on a panel gave attendees valuable insights into how Cambridge and Microsoft collaborate effectively.
Attendees took part in interactive sessions where they could try out surgical robot training tools by threading polos onto shoelaces, and visit the IMAXT virtual reality cancer lab, where they could don a VR headset and ‘fly’ inside a 3D cancer tumour to give them a completely new perspective.
Patient and student engagement
Patients spoke and participated on every panel, and the CRUK Cambridge Centre Ovarian Cancer Patient group presented a flash talk and a poster highlighting how they participate in and help shape research. The patient voice was a golden thread running throughout the event, joining leading surgeons, researchers, clinicians and industry partners at the speaker dinner at Sidney Sussex College where discussion flowed and future collaborations were planned. Patients who could not attend could access the event online and worries/questions about the use of AI in healthcare were collected from across the region and beyond from the Cambridge Cancer Research Hospital Patient Advisory Group and the Addenbrooke’s Cancer Patient Partnership Group and put to the experts at the event.
Students and early career researchers met inspirational leaders such as Dr Mahmood. Fostering new connections and enabling young Cambridge talents to connect with global experts was fantastic, with one attendee remarking on the event’s friendly atmosphere, so they had the confidence to speak to everyone, including the speakers. The feedback from attendees was very positive. One attendee said: “I particularly appreciated the involvement of patients. This aspect was unique, stimulating, and very emotional. It provided a grounding perspective for the scientists and was incredibly motivating,” while another suggested a way to build upon the successful event, commenting: “It would be great to get more clinicians, information governance and regulators involved next time.”
Preparing for the future
Following an action-packed two days, some patient queries from sessions remained so we are creating a myth-busting video about AI in healthcare to answer outstanding questions. We have filmed patients asking the questions, which will be answered by researchers and I will cover wider concerns in a podcast, again involving patients and researchers, planned for early 2025.
The symposium has had a lasting impact, with collaborations planned, funds raised and, responding to the feedback, a follow-up symposium – AI in Healthcare: Governance and Policy has just been granted funding by the Accelerate Science Programme. The potential to impact governance and policy is exciting and we are thrilled to present this next symposium on 8 July 2025 which, with our partner the Centre for Science and Policy, will bring together researchers, clinicians, patients and policy makers to discuss the shape of policy in the UK around the use of AI in a healthcare context.
This project was funded though the 2023 Accelerate-C2D3 funding call for novel applications of AI for research and innovation. You can read more about other funded projects here.
Watch the sessions from the conference here:
