Centre for Medical Image Computing (CMIC)


Konstantinos Georgiadis - Alexandra Young

Start: Oct 25, 2017 01:00 PM
End: Oct 25, 2017 02:00 PM

Location: UCL Bloomsbury - Roberts 106 Roberts building

Alexandra Young

Title - Uncovering the progression and heterogeneity of neurodegenerative diseases

Abstract - Neurodegenerative diseases are highly heterogeneous, consisting of subtypes that have different patterns of disease progression over time. Identifying these subtypes and characterising their progression is challenging, but offers the potential to provide new insights into disease mechanisms, and to enable fine-grained patient stratification for clinical trials and healthcare.

In this talk I will present Subtype and Stage Inference (SuStaIn), a machine learning technique that identifies clusters of individuals (subtypes) with common disease progression patterns. I will show results obtained from application of SuStaIn to imaging studies of two neurodegenerative diseases: genetic frontotemporal dementia and Alzheimer’s disease, which demonstrate the utility of SuStaIn for subtype discovery and patient stratification.

Konstantinos Georgiadis

Title - Computational modelling of pathogenic protein in neurodegenerative diseases

Abstract - Pathogenic protein accumulation and spread are fundamental principles of neurodegenerative diseases and ultimately account for the atrophy patterns that distinguish these diseases clinically. However, the biological mechanisms that link pathogenic proteins to specific neural network damage patterns have not been defined. We developed computational models for mechanisms of pathogenic protein accumulation, spread and toxic effects. By varying simulation parameters we assessed the effects of modelled mechanisms on network breakdown patterns. Our findings suggest that patterns of network breakdown and the convergence of patterns follow rules determined by particular protein parameters. We evaluated the predictive power of our model by comparing the order of regional volume becoming abnormal during a simulation against the output of an event based model derived from empirical data of sporadic Alzheimer's disease. We also demonstrate how our framework could evaluate candidate therapies. This work provides a basis for understanding the effects of pathogenic proteins on neural circuits and predicting progression of neurodegeneration.