Artificial intelligence (AI) could fundamentally transform clinical workflows in image-based diagnostics and population screening. AI promises more objective, accurate and effective analysis of medical images, improving the early detection of disease. Health services are overstretched and under significant pressure, so advanced technological solutions are urgently needed. However, a major hurdle in using medical imaging AI in clinical practice is the assurance whether it is safe for patients and continues to be safe after deployment. Differences in patient populations and changes in the data acquisition pose challenges to today’s AI algorithms. This Research Chair focuses on developing AI safeguards including automatic failure detection, monitoring of performance, and analysis of bias to ensure the safe and ethical use of medical imaging AI.
Recent Papers
- Kori et al. (2024) Identifiable Object-Centric Representation Learning via Probabilistic Slot Attention
- Khara et al. (2024) Generalisable deep learning method for mammographic density prediction across imaging techniques and self-reported race
- Jones et al. (2024) A causal perspective on dataset bias in machine learning for medical imaging
- Kori et al. (2024) Grounded Object-Centric Learning
- Ng et al. (2023) Prospective implementation of AI-assisted screen reading to improve early detection of breast cancer
- Roschewitz et al. (2023) Automatic correction of performance drift under acquisition shift in medical image classification
Projects
Publications
2024 (3)
- Mélanie Roschewitz, Fabio De Sousa Ribeiro, Tian Xia, Galvin Khara, Ben Glocker, Counterfactual contrastive learning: robust representations via causal image synthesis, MICCAI Workshop on Data Engineering in Medical Imaging (DEMI), 2024
- Tian Xia, Mélanie Roschewitz, Fabio De Sousa Ribeiro, Charles Jones, Ben Glocker, Mitigating attribute amplification in counterfactual image generation, International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI), 2024
- Charles Jones, Daniel C. Castro, Fabio De Sousa Ribeiro, Ozan Oktay, Melissa McCradden, Ben Glocker, A causal perspective on dataset bias in machine learning for medical imaging, Nature Machine Intelligence, 2024
2023 (5)
- Mélanie Roschewitz, Galvin Khara, Joe Yearsley, Nisha Sharma, Jonathan J. James, Éva Ambrózay, Adam Heroux, Peter Kecskemethy, Tobias Rijken, Ben Glocker, Automatic correction of performance drift under acquisition shift in medical image classification, Nature Communications, 2023
- Carolina Piçarra, Ben Glocker, Analysing race and sex bias in brain age prediction, MICCAI Workshop on Fairness of AI in Medical Imaging (FAIMI), 2023
- Mélanie Roschewitz, Ben Glocker, Distance matters for improving performance estimation under covariate shift, ICCV Workshop on Uncertainty Quantification for Computer Vision, 2023
- Charles Jones, Melanie Roschewitz, Ben Glocker, The Role of Subgroup Separability in Group-Fair Medical Image Classification, International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI), 2023
- Fabio De Sousa Ribeiro, Tian Xia, Miguel Monteiro, Nick Pawlowski, Ben Glocker, High Fidelity Image Counterfactuals with Probabilistic Causal Models, International Conference on Machine Learning (ICML), 2023