Dec - Martin McKeown¶

Speaker: Martin McKeown

Talk Title: What can computational approaches teach us about Parkinson’s Disease ?

Event Details

Date/Time:

December 11, 2008, 6:00pm

Affiliation: University of British Columbia, Department of Electrical and Computer Engineering, Biomedical Signal and Image Computing Laboratory

URL: Martin McKeown

Abstract:

Parkinson’s disease (PD) is the second most common neurodegenerative

disease in Canada, after Alzheimer’s disease. Although treatments (both

medical and surgical) are available for PD, and can have dramatic

beneficial effects especially early in the disease, they treat the

symptoms of the disease without altering the overall progression. Using

fMRI, we can non-invasively probe the normal and parkinsonian brain, but

the data require extensive processing to get meaningful results. We will

discuss the roles of Independent Component Analysis (ICA), Dynamic

Bayesian Networks (DBNs), Probabilistic Boolean Networks (PBNs), Large

Deformation Diffeomorphic Metric Mapping (LDDMM), replicator dynamics,

as well as 3D moment invariants in the analysis of these data sets.

Additionally, we will describe how second order linear dynamical system

theory can be applied to manual tracking data from PD and normal

subjects. Finally, since recent research has demonstrated that

functionally, the Parkinsonian state is characterized by the emergence

of pathological oscillations in the beta range (12-30 Hz) within basal

ganglia / cortical loops, we will describe how frequency-domain

analysis, such as partial directed coherence (PDC) can be used to

investigate the electroencephalogram (EEG) recordings from PD subjects.

These technologies will be put in the context of exploring compensatory

mechanisms in PD, capable of ameliorating overall disability.

Introductory Speaker: ### Cydney Nielsen

Talk Title: Genomic Data Visualization: Making Sense of Large-Scale Data Sets

Presentation: Download Presentation

Affiliation: Jones Laboratory, Genome Sciences Centre