An AI project from PharosAI, a collaboration between Barts Health NHS Trust, Guy’s and St Thomas’ NHS Foundation Trust, Queen Mary University of London (QMUL), and King’s College London (KCL), has received funding from Innovate UK to explore the potential to develop digital twins of human tissue.
PharosAI is focused on exploring the potential of AI in cancer research and cancer care, working on bringing together multimodal cancer data and advanced AI models, and backed by a reported £40 million in public and private investment.
The project, DESTINY, aims to understand whether AI can predict complex biological information from routinely collected tissue data, with the hopes of helping researchers better understand cancer, discover new drugs, and develop more personalised approaches to diagnosis and treatment.
According to QMUL, the funding will support an initial feasibility study to look at the technical and commercial viability of a larger-scale project. If successful, partners will be eligible to apply for up to £10 million in further funding.
Louise Jones, chief scientific officer of PharosAI, says: “DESTINY builds on PharosAI’s rich multimodal data spanning digital pathology, spatial transcriptomics, whole genome sequencing and proteomics to develop advanced AI models that can predict biological information from routine tissue samples. These digital twins of human tissue could enrich and expand the datasets available for research and, ultimately, transform how molecular analysis is carried out in clinical settings.”
DESTINY will develop advanced AI models and federated leaning methods to promote learning from larger and more diverse datasets. QMUL highlights the use of a federated learning approach to tackle challenges around privacy, security, and governance when using biomedical data for research purposes, allowing AI models to learn from data held across different organisations without bringing it into a central location.
“The team will test whether these approaches can predict molecular information that is difficult or expensive to measure directly, using routinely collected tissue data,” QMUL continues. “In the long term, the team aims to use these models to create ‘digital twins’ – digital representations of human tissue that predict information about its biological characteristics.”
Wider trend: Innovation in cancer care
A partnership has been launched between the DATA-CAN data research hub for cancer hosted at Leeds Teaching Hospitals and the West Midlands Secure Data Environment to support access to high quality cancer data for researchers. The partnership draws on pseudonymised patient data from patient records at major cancer centres across the UK, including on diagnosis, treatments, test results, and patient outcomes. The first dataset has been uploaded to the WM SDE, with three further datasets currently in development, Leeds Teaching shares.
AI enabled breast cancer screening is reportedly helping patients at Harrogate and District NHS Foundation Trust (HDFT), thanks to the introduction of a new scanner funded by £40,000 from the Harrogate Hospital & Community Charity. The AI enabled HER2 tissue scanning system works by capturing high resolution digital images of breast cancer tissue samples, using an AI algorithm to help pathologists assess HER2 status by detecting the presence of the HER2/neu gene that can be a cause of breast cancer, the trust shares.
An NHS cancer innovation programme has been launched with the aim of improving productivity, managing demand or capacity in cancer care pathways, and supporting earlier detection and diagnosis of cancer through screening or diagnostic tools. Two funding models are offered to support varying stages of innovation maturity and overcome barriers to adoption, with the first offering up to £3 million over a maximum of 24 months for single UK or EU organisations with “demonstrable partnerships” with NHS organisations in England and an independent evaluation partner. The second model is open to ICBs or cancer alliances based in England as the lead applicant, offering up to £5 million over the first 24 months of a project lasting a maximum of 36 months.
