Date of Award


Document Type

Open Access Master's Report

Degree Name

Master of Science in Health Informatics (MS)

Administrative Home Department

Department of Applied Computing

Advisor 1

Weihua Zhou

Committee Member 1

Guy Hembroff

Committee Member 2

Donald Peck


The heterogeneous nature of today’s evolving health databases requires new techniques and approaches to process these data and extract clinically useful information. This relevant information obtained can be used to improve the response rate of cardiac resynchronization therapy (CRT) in patients with heart failure. Hierarchical clustering (HC) which is an unsupervised ML technique may uncover clusters within the bulk of data of patient population which is useful for strategies towards precision and personalized medicine. This study aims to investigate how HC can be used to automatically group a bulk of clinically acquired CRT data into clusters and subgroups that could confer clinically relevant information. About 165 patient data were used in the study and the analysis resulted in 4 different phenogroups with varying response rates. Some features were statistically significant when compared within the subgroups. Lastly, the study concludes that HC can be used to integrate and analyze different kinds of clinical data to aid in the identification of HF patient subgroups that are likely to respond to CRT.