Weekly seminar: From the scanner to the cap, explainable deep learning tools for cognition in ADHD
Speakers
Dr Jack Scott
Topic
In this talk I will discuss some of our recent findings from the Human Affective & Motivational Neuroscience lab at Otago. I will talk about using deep learning for predicting cognitive functioning from functional MRI timeseries, and how we extended these tools to high-density electroencephalography – The electrical rhythms of the brain. I will then discuss my fellowship project. This project involves building on these tools (with some cool tricks) to create a measurement tool that can accurately detect medication-related changes in brain rhythms in ADHD kids in Aotearoa NZ. The goal is a tool that can tell us how well ADHD medication is working. I will share some of our promising early preliminary data, and how we aim to make the tool explainable, unbiased, and hopefully fit for purpose in the future as a support tool for clinicians.
About Speaker
Jack is a Postdoctoral Fellow at the University of Otago, under Dr. Narun Pat. Jack previously spent most of his 20s as a support worker for kids and young adults with neurodevelopmental conditions, before returning to University in 2017 to try to get a bachelors degree. Falling in love with neuroscience in the process, he went on to do a PhD looking at schizophrenia, and after finishing his PhD transitioned to neuroimaging, ADHD and AI research. In his spare time he has recently started getting into DIY and can sometimes be found window shopping the power tools section at Bunnings.
