UCL Centre for Medical Image Computing


Baris Kanber & Ahmed Karam - CMIC/WEISS Joint Seminar Series

26 January 2022, 1:00 pm–2:00 pm

Baris Kanber & Ahmed Karam- talks as part of CMIC/WEISS Joint Seminar Series

Event Information

Open to





UCL Centre for Medical Image Computing and Wellcome/EPSRC Centre for Interventional and Surgical Sciences

Speaker: Baris Kanber

Title: Big data/machine learning analysis of clinical MRI scans in multiple sclerosis - the experience from our imaging response to treatment study

Machine learning and big data may empower the development of precision medicine approaches in the treatment of multiple sclerosis, as well as other health conditions. In this talk, I will be presenting our recent, big data/machine learning work on predicting the imaging response to treatment in MS using routinely acquired MRI data and share the difficulties we have encountered along the way. Published as “High-dimensional detection of imaging response to treatment in multiple sclerosis” (npj Digital Medicine 2019; 2(49)), the work combined contemporary machine learning techniques and routinely acquired brain MRI data to study the imaging changes that occur in the brain in response to a disease modifying therapy. This has partly formed the basis for our ongoing work in the development of personalised medicine approaches in the treatment of multiple sclerosis. Briefly, we were able to construct fingerprints of change that occur within the brain before and after treatment and show that the imaging response to treatment can be captured more sensitively with high-dimensional methods than with the conventional radiological measures (AUC: 0.890 [95% confidence interval: 0.885-0.895] vs. 0.686 [95% confidence interval: 0.679-0.693]). The talk would be of interest to those planning to use routinely acquired clinical imaging data in a big-data setting where confounders are aplenty in combination with machine learning which comes with its own pitfalls.

Speaker: Ahmed Karam

Title: Bayesian Bacterial Detection using Irregularly Sampled Optical Endomicroscopy Images

Pneumonia is a major cause of morbidity and mortality of patients in intensive care. Rapid determination of the presence of the pathogenic bacteria in the distal lung may enable a more tailored treatment regime. Optical Endomicroscopy (OEM) is an emerging medical imaging platform with preclinical and clinical utility. Pulmonary OEM via multi-core fibre bundles has the potential to provide in vivo, in situ, fluorescent molecular signatures of the causes of infection and inflammation. In this seminar, I will present a Bayesian approach for bacterial detection in OEM images. The model considered assumes that the observed pixel fluorescence is a linear combination of the actual intensity value associated with tissues or background, corrupted by additive Gaussian noise and potentially by an additional sparse outlier term modelling anomalies (bacteria). The bacteria detection problem is formulated in a Bayesian framework and prior distributions are assigned to the unknown model parameters. A Markov chain Monte Carlo algorithm is used to sample the posterior distribution of the unknown parameters. Extensive simulations conducted using synthetic and real data will be presented.

Chair: Danny Alexander