Electronic Theses and Dissertations

Date

2025

Document Type

Dissertation

Degree Name

Doctor of Philosophy

Department

Biomedical Engineering

Committee Chair

Aaryani Tipirneni-Sajja

Committee Member

Cara Morin

Committee Member

Carl D. Herickhoff

Committee Member

Deepak Venugopal

Abstract

Hepatic steatosis and iron overload are common manifestations of diffuse liver disease. They can cause lipotoxicity and iron toxicity respectively via oxidative hepatocellular injury and can lead to progressive fibrosis, cirrhosis, and eventually, liver failure. Over the past two decades, magnetic resonance imaging (MRI) has emerged as a non-invasive tool to diagnose steatosis and iron overload. Multi-spectral fat-water models incorporating R2* correction have been proposed to simultaneously quantify fat and iron overload. However, there is still an ongoing debate on whether the multi-spectral signal model assuming same R2* (single-R2*) or different R2* (dual-R2*) for fat and water is accurate in the presence of fat. Additionally, current studies lack thorough investigation of R2* techniques for fat and iron quantification covering the entire clinical range. Apart from this, clinical reporting of steatosis and iron overload from MRI images requires manual liver segmentation, which is time intensive and can have reader bias, hence serving as a bottleneck in the clinical workflow. To overcome these limitations, the purpose of this dissertation is firstly, to evaluate the performance of multispectral fat-water models for fat and iron quantification using simulations covering the entire clinical spectrum and validating using phantoms and secondly, to design an automatic liver segmentation algorithm for expediting the clinical reporting of hepatic iron overload.

Comments

Data is provided by the student.

Library Comment

PDF

Notes

Embargoed until 10-04-2025

Download

Available for download on Saturday, October 04, 2025

Share

COinS