Assistant Professor of Research Neurology
University of Southern California
I’m interested in understanding how neurodegenerative diseases such as Alzheimer’s and Parkinson’s progress over time.
My research involves modelling the evolution of these diseases by combining computational modelling, statistics and machine learning with data from neuroimaging, cognitive testing, biofluids such as cerebrospinal fluid (CSF) and other complementary data.
I have worked on:
Applying disease progression models, such as Subtype and Stage Inference (SuStaIn; see Alex Young’s paper) to model the spread of Alzheimer’s pathologies within the brain. Our paper applied SuStaIn to (in vivo) amyloid PET, tau PET and (post mortem) neuropathology based measures to show that the progression of amyloid protein and tau protein based pathologies is best described by two subtypes. We found an ‘amyloid-first’ subtype in which amyloid pathology spreads throughout the brain prior to the spread of tau and a ‘tau-first’ subtype in which tau pathology precedes amyloid. The ‘amyloid-first’ subtype reflects the typical course of Alzheimer’s disease while the ‘tau-first’ subtype is heterogeneous. Tau-first can be further stratified by genetic risk of Alzheimer’s disease (APOE4 carriage): those who are early tau-first with a lower genetic risk (APOE4 negative) are unlikely to accumulate amyloid and therefore probably do not belong on the Alzheimer’s spectrum. Those who are early ‘tau-first’ with higher genetic risk (APOE4 positive) are rare but may belong within the Alzheimer’s spectrum due to their increased rate of amyloid accumulation.
Co-developing (along with Peter Wijeratne) the pySuStaIn package, a Python implementation of SuStaIn. Check out our software paper.
Developing probabilistic biomarker trajectory models to characterize the course of neurodegeneration (paper)(abstract). The toolbox is here.
Developing longitudinal pattern recognition and machine learning based models to discriminate early Alzheimer’s (paper) and Parkinson’s disease (thesis). You can find the toolbox here.
I’m also interested in making easy-to-use tools for visualizing results in neuroimaging.
A full list of my publications is here.
I am currently an assistant professor at USC, where I work on disease progression modeling of neurodegenerative diseases. Previously I was a postdoctoral research associate at UCL from 2017 to 2020. I was a proud member of both the POND group run by Danny Alexander and Neil Oxtoby and the COMBINE lab run by Andre Altmann. Prior to UCL I completed my PhD in neuroimaging at King’s College London in early 2017, working with Andre Marquand and David Lythgoe on machine learning based methods that discriminate early neurodegeneration.
I spent six years in quantitative finance as a programmer/analyst/trader. Prior to that I completed a Master’s in aerospace engineering at the University of Maryland, working on robotic manipulator control algorithms and some motor control electronics. Prior to that I studied at Johns Hopkins University, majoring in electrical engineering.