Analysis and Visualization Methods for Data-Driven Longitudinal Patient Summary

Public Deposited
Analytics

Downloadable Content

Download PDF
Request Version for Screen Reader
Last Modified
  • February 28, 2019
Creator
  • Krishna Sastrula, Jyotsna
    • Affiliation: School of Information and Library Science
Abstract
  • Digitization of health records has opened avenues for intensive research in the fields of health informatics. Power of machine learning, statistical analysis and visual analytics could be utilized to make optimal use of this information. The proposed project is to develop an interactive visualization tool that summarizes a patient's medical history, highlighting all his/her important events based on the knowledge of similar patients. Given a set of patients with common conditions, statistical analysis can be used to develop models that prioritize features based on associations between features and condition-specific outcome measures. This manuscript in particular describes the model developed to prioritize a patient's events from his medical history. The model is trained with the population of patients and their events. Their correlations with the outcome variable are calculated to identify the important events in a specific cohort. This correlation score can be used to prioritize the events associated with an individual patient. This model is one of the models that will be used to summarize an individual patient's medical data via interactive visualization methods.
Date of publication
Subject
DOI
Resource type
Rights statement
  • In Copyright
Advisor
  • Gotz, David
Degree
  • Master of Science in Information Science
Degree granting institution
  • University of North Carolina at Chapel Hill
Extent
  • 35
Deposit record
  • 93b21495-b8dd-4a0b-8039-3577fd40d51d

Relations

Parents:

This work has no parents.

Items

Thumbnail Title Date Uploaded Visibility Actions
file details jyotsnakrishnasastrula.pdf 2019-05-13 Public Download