Vaishnavi Kulwal
BME MS Thesis Defense Presentation
Date: 2025-07-14
Time: 1:00- 2:30 pm
Location / Meeting Link: HSRBII N100/ https://gatech.zoom.us/j/98673538654
Committee Members:
Dr. David A. Reiter; Dr. Candace C. Fleischer; Dr. John Oshinski
Title: Multi-Tissue IVIM MRI Measurement of Perfusion in Diabetic Foot Ulceration with Semi-Supervised Bone Segmentation
Abstract:
Diabetic foot ulcers (DFU) are a serious complication of type 2 diabetes mellitus (T2DM). Prolonged hyperglycemia in T2DM patients causes blood perfusion complications due to calcification of blood vessels and peripheral neuropathy, leading to poor wound healing. Existing clinical tools for assessing perfusion in DFU often lack tissue specificity and have limited prognostic value and do not measure at a microvascular level. In contrast, MRI-based methods like intravoxel incoherent motion (IVIM) imaging offer a non-invasive, microvascular level approach to measure tissue-level perfusion. Previous work has demonstrated the feasibility of IVIM for detecting perfusion differences in the foot compartments differentiated on the basis of blood source artery, across the diabetes progression spectrum. Building on that, the first aim of the current study focuses on tissue-level perfusion differences among patients at different levels of disease progression namely: diabetic patients with active foot ulcers, diabetic patients without foot ulcers and controlled diabetes, and healthy controls. Perfusion characteristics were evaluated within key tissue compartments, mainly bone, fat pad, and muscle to investigate whether microvascular dysfunction can be differentiated across tissue types. The second aim of this study reports the feasibility of an automated segmentation pipeline, using a human-in-the-loop approach for high throughput tissue segmentation. The bone tissue is used to evaluate the performance of these pipeline-generated segmentations. The results showed good agreement with expert manual annotations. These findings support the potential of IVIM MRI as a valuable, tissue-level perfusion imaging modality for DFU characterization, while also highlighting the viability of automation-assisted tissue segmentation for robust, scalable analysis.