Data collection, curation and annotation are critical components in machine learning for medical image analysis which can have profound effects on the overall performance, robustness and generalisation. In this talk, we will illustrate this problem with results from an empirical study using multi-site neuroimaging data. We will discuss mechanisms for assessing the readiness of clinical data for machine learning and we will consider methods for pre- and post-analysis quality control. We conclude by discussing how causal reasoning may help us to identify data issues early on, and find solutions to tackle the key challenges of domain shift and data scarcity.
0.5 CPD credit
Watch the video and complete the online self-reflection form. Go to "My events" to download your certificate. You must login to watch the video and receive CPD.
Ben Glocker is Reader in Machine Learning for Imaging at the Department of Computing at Imperial College London where he co-leads the Biomedical Image Analysis Group. He also leads the HeartFlow-Imperial Research Team and is scientific advisor for Kheiron Medical Technologies and a Visiting Researcher at Microsoft Research Cambridge. He holds a PhD from TU Munich and was a post-doc at Microsoft and a Research Fellow at the University of Cambridge. His research is at the intersection of medical image analysis and artificial intelligence aiming to build computational tools for improving image-based detection and diagnosis of disease.