Mitral valve regurgitation is the second most common clinically significant valvular defect and is a common degenerative condition that occurs with aging. Prevalence varies from around 3% in children to around 80% in those over age 80. Other than age related degeneration, there are a variety of diseases such as Marfan Syndrome, endocarditis, lupus, Down’s Syndrome, rheumatic fever, and cardiac ischemia that cause mitral valve disease. Valve mechanics are highly importance in the development and progression of disease. We are specifically interested in comparing strain profiles between patient groups which would allow us to investigate how tissue in various disease states changes mechanically as well as how mechanical changes in the leaflet can impact valve dysfunction and disease. Strain profiling also offers diagnostic potential: from echocardiogram alone it is impossible to determine specific disease etiologies of the mitral valve. If a mechanical pattern can be determined to underlie individual valvular pathologies, it can be used as a specific diagnostic.
Our lab has developed a MATLAB program that is able to map and quantify strain profiles of mitral valve leaflets from patient echocardiograms and MRI. This program has demonstrated its ability, in a preliminary experiment using a cohort of Weill Cornell Medical Center patients, to capture strain variances across the cardiac cycle, and its ability to differentiate between and identify differences between anterior and posterior leaflets in the mitral valve. Despite using very coarse spatial modeling, it demonstrated the ability to capture mechanical information between the anterior and posterior leaflets supporting methodological utility, especially with improvements upon the software. Currently, this program requires extensive pre-processing making it not particularly user friendly, is slow to run due to it being such a large, complex program run in MATLAB, and does not deliver as reliable and sensitive data as it should. We are looking for someone to better automate the process and take it out of MATLAB into a more global, user friendly language. With these improvements, we foresee this program being translated into a more widely commercially viable implementation and being made available as an add on for echocardiogram software, allowing for real time strain mapping, and preemptive diagnoses at time of the scan.
Advisor Name: Jonathan Butcher
- Email - jtb47@cornell.edu
Co-Advisor Name: Jason Spector
- Email – jas2037@med.cornell.edu
Interested candidates should email a statement of interest and suitability of background/training, an updated resume, and a fall semester class schedule to Jonathan Butcher (jtb47@cornell.edu). Teams of students are also accepted.