Medicine & Dentistry-Dentistry and Dental Hygiene
University of Alberta
Summary of research:
High-dimensional data arises from a variety of medical imaging devices, sensor networks, and genomic sequencing technology. This data is able to describe a single feature from multiple perspectives and as such, the quantity of data for analysis of becomes extremely large and complex very quickly. Persistent homology is a new method of managing high-dimensional data sets by studying shapes and spaces. Persistent homology techniques are believed to be superior to traditional methods for multidimensional data; they can effectively identify differences between datasets. Functional data analysis analyzes variations between data elements. Taking advantage of both approaches, my research program focuses on the development of sophisticated statistical tools to comprehensively compare medical images of patients. Since my research involves the study of routine images captured by common three-dimensional imaging devices (including those used by sleep medicine specialists) my findings are transferrable to the clinical setting. The proposed study will develop statistical tools to match the faces of patients with the ideal facemask for their night time CPAP device in order to treat sleep apnea. Our tools will be used by health care practitioners to maximize patients' use their CPAP devices. In turn, this will improve the patient's treatment and health outcomes in both the short and the long-term.