I'm now pursuing my master's degree in AI/ML applied research at University of Toronto starting from September 2020. I expect to graduate in December 2021.
Developing Neural Processing Unit software.
Research under Professor Marzyeh Ghassemi's supervision focusing on medical radiology image diagnosis. Working on radiology report generation / disease classification
Develop Remote Evaluation web application for High-Speed Converter. Including backend in Python using Django and Celery and control over Spectrum Analyzer and Signal Generator using SCPI.
External spokesperson of UTRT and regularly interact with other student organizations and university officials as well as overseeing the process of student organization event planning and giving deliberate advice to event department.
cGPA 3.89/4.0
Guanxiong Liu*, Tzu-Ming Harry Hsu*, Matthew McDermott, Willie Boag, Wei-Hung Weng, Peter Szolovits, Marzyeh Ghassemi
In this work, we present a domain-aware automatic chest X-Ray radiology report generation system which first predicts what topics will be discussed in the report, then conditionally generates sentences corresponding to these topics. The resulting system is fine-tuned using reinforcement learning, considering both readability and clinical accuracy, as assessed by the proposed Clinically Coherent Reward.
Accepted by Machine Learning for Healthcare, 2019
Preprint available on arXiv
Spotlight Talk available Here at 01:01:23
Laleh Seyyed-Kalantari, Guanxiong Liu, Matthew McDermott, Marzyeh Ghassemi
In this work, We demonstrate that TPR disparities exist in the state-of-the-art classifiers in all datasets, for all clinical tasks, and all subgroups. We find that TPR disparities are most commonly not significantly correlated with a subgroup's proportional disease burden; further, we find that some subgroups and subsection of the population are chronically underdiagnosed. Such performance disparities have real consequences as models move from papers to products, and should be carefully audited prior to deployment.
Preprint available on arXiv