Search results

(1 - 1 of 1)
A KINETIC MODEL TO ESTIMATE VASCULAR PERMEABILITY FROM FLUORESCEIN VIDEOANGIOGRAPHY DATA
Title
A KINETIC MODEL TO ESTIMATE VASCULAR PERMEABILITY FROM FLUORESCEIN VIDEOANGIOGRAPHY DATA
Creator
Hu, Shaoxian
Date
2017, 2017-05
Description
Diabetic retinopathy (DR) is a disease affecting the tissue of the retina through hemodynamics and vascular damage that is a side effect of...
Show moreDiabetic retinopathy (DR) is a disease affecting the tissue of the retina through hemodynamics and vascular damage that is a side effect of hyperglycemia. As such, markers of hemodynamic and vascular irregularity have been proposed as potential early indicators of the disease. However, conventional approaches to estimate these conditions are not sensitive enough, leading to indications that are too late, appearing only after irreversible vision problems have occurred. Indicator-dye-dilution tracer kinetic modeling is a decades-old methodology that can be used to quantify blood flow and vascular permeability based on the kinetics of a rapidly injected blood pool imaging agent. The methodology is able to track sensitive hemodynamic changes in stroke and heart disease patients, yet it has never been applied to blood flow and vascular permeability mapping of the retina. The purpose of my Master’s thesis was to adapt those tracer kinetic methods to fluorescein retinal videoangiography data collected in human to map both blood flow and vascular permeability so that in future work, changes in these parameters can be quantified at varying stages of the development of DR. Studies were carried out on simulated data, healthy and diabetic rats, and healthy and diabetic humans. While a robust validation of measures of blood flow and vascular permeability are still required, this first-of-its-kind study exposed many unique complications in applying the kinetic models to human fluorescein videoangiography data. Specifically, eye motion must be accounted for, both detector and fluorescein fluorescence quenching signal saturation must be avoided, and standard use of automatic gain control of imaging exposure should not be used (though corrections were developed for this); rather, data should be collected at a stable exposure throughout a measurement. Solutions were developed for all of these complications that were tested in rat models (optimal data collection in humans is ongoing). In rat study, average vascular permeability in the rat measured by the “extraction fraction”, the extraction fraction in control subject (n=3) 9.4e-4±1.3e-3, is lower than STZ-diabetic subjects (n=3) 5.5-2±5.17e-2 and VEGF group (n=3) 7.6e-2±5.1e-2. In healthy human volunteers, retinal extraction fraction was measured to be 5.0e-5 ± 6.1e-5 (n=2), whereas this was found to be elevated to 1.2e-3 ± 1.0e-3 in patients for diagnosed DR (n=3). These early results clearly demonstrated a statistically significant elevation in vascular permeability in all diabetic groups compared to controls (p<0.05). The human study was carried out on advanced cases of patients who were known to have DR as a test case; however, the sensitivity of the approaches is demonstrated by the eatly changes in extraction fraction observed at only 48 h after the onset of hyperglycemia in the rat experiments (well before irreversible damage to the retina had occurred). These results are promising for the adaptation of our modified kinetic modeling approach to measure hemodynamics and vascular permeability changes during the early onset of DR when preventative therapies could be most effective.
M.S. in Biomedical Engineering, May 2017
Show less