Icon Legend

This session is not in your schedule.

This session is in your schedule. Click again to remove it.

Biomedical Imaging and Instrumentation

Biomedical Imaging and Instrumentation - Poster Session C & D

U45 - Computational model to diagnose RHD from non-doppler echocardiography

Friday, October 10, 2025

10:00 AM - 11:00 AM PDT

Location: Exhibit Hall F, G & H

Presenting Author(s)

BE

Blake Evans

Graduate Student
The University of Texas at Austin
Austin, Texas, United States

Co-Author(s)

SN

Shreya Nallapa

Research Assistant
The University of Texas at Austin, United States
EA

Estrella Ayala Hernandez

Undergraduate Research Assistant
The University of Texas at Austin
Austin, Texas, United States

Last Author(s)

Edward Castillo, PhD

Associate Professor & Associate Chair
The University of Texas at Austin
Austin, Texas, United States

Introduction: : Rheumatic heart disease (RHD) is a cardiac disease caused by rheumatic fever, a condition that most commonly stems from untreated group A streptococcal (GAS), commonly known as strep throat [1]. Without proper diagnosis, RHD can lead to heart failure and death. Very few cases of RHD occur in the United States and similar developed nations as readily available antibiotics prevent strep throat from progressing to rheumatic fever. Further, isolated cases are likely to be diagnosed early via doppler echocardiography, as mitral valve regurgitation is a key diagnostic marker for RHD, leading to improved long-term patient outcomes. However, RHD continues to affect more than 33 million people globally and is endemic in low- and middle-income countries (LMICs) in Africa, Asia, and South America [2]. Despite efforts to improve RHD diagnosis in these regions, approximately 15% of adults with RHD in LMICs die within three years, a number that likely underestimates total deaths as many patients die without a diagnosis. The lack of access to doppler echocardiography and skilled technicians is a significant barrier to early RHD diagnosis in LMICs. RHD tends to be asymptomatic, aside from the preceding strep throat, and the first cardiac symptoms can be fatal. To this end, we present a preliminary computational framework to automatically quantify mitral valve geometry and velocity to effectively and reproducibly quantify RHD diagnostic markers and screen patients for early-stage RHD.

Materials and

Methods: : This is a retrospective study using non-doppler apical four chamber (A4C) echocardiograms from the Ascension Seaton Medical Center in Austin, TX (IRB: STUDY00005626). Under the IRB protocol data cannot be shared publicly. The dataset for this preliminary study consists of two positive RHD patients (as identified by doppler echocardiography) and one healthy control.
First, the anterior and posterior leaflets of the mitral valve were segmented using ITK-snap. Due to the relatively low signal-to-noise ratio and two-dimensional nature of echocardiography, the full mitral valve is not visible on all frames of each echocardiogram. In these cases, the segmentation mask is linearly interpolated from neighboring frames in the video.
In an echocardiogram, the mitral valve resembles an ellipse-like shape. To quantify this geometry, we identify the mid-line that constitutes the path along the long axis of the valve, on the anterior and posterior side respectively. By integrating orthogonal segments to this line, we calculate the total surface area of the valve. Additionally, we calculate length, width, and aspect ratio.
As the valve pivots from the basal rim of the myocardium, we quantified radial, angular, and tangential velocity of the anterior and posterior mitral valve. To do so, we performed linear regression along the midline of the axis, to find the straight line that best represents the current angle and length of the respective valve to represent the valvular motion in polar coordinates, eliminating the effect of wall contraction and overall heart motion from our velocity calculation.

Results, Conclusions, and Discussions:: Results
Our most interesting result is the difference in total angular deformation of the posterior leaflet of the mitral valve in the healthy patient compared to those with RHD. The healthy patient’s posterior mitral leaflet rotated roughly 30° during diastole, while the patients with RHD both rotated less than 20°. It can visually be seen that the RHD patients’ mitral valve did not fully close during systole. Additionally, the angular velocity of the posterior leaflet was 25% slower on average for the RHD patients compared to the healthy patient. Lastly, the motion of the anterior mitral valve leaflet was similar among all three patients.
Discussion
Mitral regurgitation identified in doppler echocardiography primarily stems from mitral stenosis (MS), the narrowing of the gap between leaflets during diastole [3]. MS occurs when the mitral valve leaflets thicken and deform from their healthy state, and we postulate that identifying this deformation would be an ideal screening tool for RHD when doppler echocardiography is not available. Based on our preliminary results, we are able to successfully quantify markers for MS without doppler echocardiography.
Conclusion
This work presents a potential screening tool to identify early stage RHD in LMICs. However, the study size was very small, and we must expand our cohort to test the generalizability of our model. Next steps will focus on automatic segmentation of the mitral valve and expansion to additional echocardiographic views to capture the three-dimensional nature of the left ventricle and mitral valve. We will also go beyond a linear approximation of the valve, and use cubic splines to capture the change in curvature of the valve throughout the systolic cycle.

Acknowledgements and/or References (Optional): : 1. Kumar RK, Antunes MJ, Beaton A, et al. Contemporary Diagnosis and Management of Rheumatic Heart Disease: Implications for Closing the Gap: A Scientific Statement From the American Heart Association. Circulation. 2020;142(20):e337-e357. doi:10.1161/CIR.0000000000000921
2. Karthikeyan G, Ntsekhe M, Islam S, et al. Mortality and Morbidity in Adults With Rheumatic Heart Disease. JAMA. 2024;332(2):133-140. doi:10.1001/jama.2024.8258
3. Korzan S, Jones E, Mutneja R, Grover P. Mitral stenosis due to rheumatic heart disease - A rare cause of massive hemoptysis. Respir Med Case Rep. 2018;24:35-39. doi:10.1016/j.rmcr.2018.03.007