The Upper Gastrointestinal (UGI) multidisciplinary team (MDT) makes critical treatment decisions in every oesophageal cancer (OC) patient’s journey (e.g. surgery or chemotherapy). This has a profound impact on survival and quality of life.
Recent studies have highlighted that rising caseloads and time pressures can lead to variable and sometimes suboptimal decisions. In this context, machine learning (ML) approaches offer the potential to standardize and produce consistent, data-driven decisions. Mirroring the decision-making process by incorporating critical variables chosen by humans, will ensure that the solution is transparent, trustworthy, and ultimately, ensure routine clinical use to improve OC patients’ lives. This pump-priming award will enable the collection, processing, curation, and interpretation of image-based data (CT-scan and pathology slides) for later incorporation into a bespoke ML model. Explainability and trustworthiness of the model will be enhanced by the application of spatial transcriptomics to biopsy slides to discover the “ground truth” of gene expression programmes in cellular neighbourhoods that underpin model performance. This award will facilitate accelerated delivery of a comprehensive project to deliver “A Trustworthy Mirrored Machine Learning approach to predict Upper Gastrointestinal Multidisciplinary team treatment decisions in Oesophageal Cancer”.
