HDR UK Applied Analytics Seminar led by Dr Grant Mair, Dr Wenwen Li, and Alessandro Fontanella
As part of the HDR UK Applied Analytics scientific theme this seminar on The Challenges and Applications of Deep Learning Methods on Medical Imaging for Acute Ischemic Stroke led by Grant Mair, Wenwen Li, and Alessandro Fontanella will take place on Wednesday 19th January 2022 from 11:00 - 12:00 UK time.
This seminar will be recorded
With huge amounts of brain CT data generated daily in routine clinical practice, there is great potential to harness the predictive power of machine learning (ML) and especially deep learning (DL). However, developing DL methods for stroke, especially acute ischemic stroke (AIS) is challenging for several reasons. First, access to imaging data for research is rightly restricted, rarely centrally stored, or standardised. Second, gold-standard expert human interpretation of the imaging is qualitatively assessed by radiologists, highly variable and recorded in free text, while a quantitative ground-truth is rarely achievable. Third, standardising clinical brain CT data for DL in terms of quality and format is a highly time-intensive process with very limited prior research specific for AIS, and no fully open-source code available. Fourth, the black-box nature of DL makes the identification and evaluation of reliable biomarkers more problematic than expected. In this talk, we will show the gap between routine clinical CT image quality and the data quality requirements for ML or DL. We will describe our bespoke pipeline for processing routinely-acquired clinical CT for DL methods from a large international trial dataset – The Third International Stroke Trial (IST-3). Then we will present our novel DL method for learning AIS lesion patterns on CT and discuss its efficacy and clinically relevant observations. Finally, we will share examples of our attempts to visualise the regions of interest detected by our DL method as it attempts to identify stroke lesions and explore the potential and challenges of making meaningful interpretations from DL. (Dr Grant Mair)
Dr Grant Mair
I am a Senior Clinical Lecturer and Honorary Consultant Neuroradiologist and split my time between academic work at the University of Edinburgh and clinical work with NHS Lothian.
My primary research focus is the imaging of stroke, especially seeking relationships between imaging biomarkers and response to treatment. I am exploring automated computational methods for the assessment or imaging data and imaging metadata including radiological reports. I am particularly interested in maximising the benefit of baseline non-enhanced CT which is usually acquired first line in patients with symptoms of stroke.
Dr Wenwen Li
Dr Wenwen Li is working as a Research Fellow in a machine learning project for acute ischemic stroke lesion at Centre for Clinical Brain Sciences the University of Edinburgh. She received the B.Eng. degree in software engineering from Anhui University, China in 2008, the M.Sc. (with distinction) in operational research from The University of Edinburgh, U.K. in 2013 and PhD in Computer Science from University of Nottingham, 2018. Her current research interests include deep learning, machine learning and data analysis with the applications to medical imaging.
Alessandro Fontanella
I’m a PhD student in Biomedical AI at the university of Edinburgh. My research revolves around deep learning methods for stroke detection and other biomedical applications (such as retinal imaging). I’m also interested in generative models for counterfactual generation to improve model interpretability and performance on classification tasks.