Home Department: Engineering
Mentor: Euan Ashley (Medicine - Cardiovascular Medicine, Genetics, and Biomedical Data Science)

2022 Research Project:“Development of a Polygenic Risk Score of the Left Ventricular End Diastolic Volume and Mass using a Deep Learning Approach”

An accurate understanding of how the genome contributes to observable characteristics is becoming crucial for personalized therapy in cardiovascular medicine. Cardiomyopathies are known to have a strong genetic background. Shriya’s project aims to develop a polygenic risk score of left ventricular end diastolic volume and mass, using Magnetic Resonance images and whole-genome sequencing data from the UK Biobank. Firstly, a machine learning model will be trained to automatically label the MRI dataset. Secondly, the obtained volume and mass data will be used for the genome-wide association (GWA) study, using a machine learning model. Lastly, a polygenic risk score will be developed using Stanford University patient data. Cardiomyopathy risk prediction, contingent on an individual’s genomic data, is a powerful tool to develop for both preventative medicine and personalized treatment.
 

2024 Research Project:"Detecting Hypertrophic Cardiomyopathy Through Whole Genome Sequencing and Deep Phenotyping"

Shriya will be creating a model to predict the risk of hypertrophic cardiomyopathy (HCM) using deep learning tools on genomic, phenotypic, and imaging data available in the UK Biobank. She will develop and validate transformer models, which emphasize where a physician’s attention is drawn to in medical images, to extract dozens of cardiac measurements that are clinically relevant for predicting HCM. These parameters will be combined with whole genome sequence data to produce an HCM risk score for each patient. Her research will contribute to more precise genetic risk prediction for HCM and lay the groundwork to better treat HCM patients with gene editing modalities.