Developing intelligent image-based predictors of GI & lung cancer risk
- Challenge: Applying computer vision to medical imaging
Although medical imaging has transformed practice, the analysis of images has hardly changed and is still largely qualitative. Oxford researchers are applying machine learning and AI methods developed from multiple externally funded programmes to simplify image acquisition (endoscopy, ultrasound, CT, MRI, histopathology) to be assessable by generalists, transforming patient pathways. Oxford researchers will also provide quantitative evaluations, including risk scores. For example, in endoscopy developing novel real-time 3D imaging approaches for detecting and monitoring Barrett’s oesophagus.