Year: 2025
Type: dissertation
Abstract: This thesis details the development and evaluation of a GenAI-powered Synthetic Data Toolbox for Health Science. The primary aim of this work is to document the process of developing the toolbox which aims to mitigate data access delays in health research by providing a mechanism to generate usable synthetic health data, thereby accelerating the preliminary testing phase of health research. Key objectives include the development and integration of specialized generative models for unstructured synthetic health data, specifically ECG signals, and the implementation and evaluation of a proof-of-concept system using large language models to automatically extract parameters from uploaded literature. The thesis focuses on integrating these capabilities into a functional, open-access toolbox with a user-friendly inter-face. The central research question explores the application of generative AI methods to produce plausible and useful synthetic ECG and tabular patient datasets for health research. The methodology adopts a constructive research approach, emphasizing solving a practical problem along with academic contribution. This thesis demonstrates the feasibility of generating synthetic data for pipeline testing within health sciences and showcasing the potential of generative AI to address critical data accessibility challenges.
Author: Ranta, Janika
Institution: Tampere University
Language: English
Subfield: Health Informatics
Field: Medicine
Domain: Health Sciences
Open Access status: green
OpenAlex ID: W7160977102