Deep video networks estimate left ventricular ejection fraction (EF) from echocardiograms with expert-level accuracy, but the compute cost of running them is rarely reported. This leaves anyone building a handheld or bedside tool without clear guidance on what to deploy. We measured the accuracy-versus-compute tradeoff for EF estimation on EchoNet-Dynamic by training 22 configurations that varied clip length (8 to 64 frames), frame sampling period (1 to 4), and backbone: R(2+1)D-18, R3D-18, MC3-18, X3D-S, X3D-M, and a 2D ResNet-18 with temporal pooling. All models used one fixed training recipe. Every configuration was evaluated for accuracy using mean absolute error, R-squared, and Bland-Altman agreement; clinical utility using sensitivity and specificity at the clinically relevant EF cutoffs of 40% and 50%, plus error stratified by EF band; and cost using floating-point operations, parameter count, GPU and CPU latency, and peak memory under a single frozen measurement protocol. We stress-tested the main findings with replicate training seeds. Sparse temporal sampling outperformed dense sampling at matched frame budgets. A sampling period of 4 outperformed a period of 1 at every tested frame count while also reducing per-video cost. In the seed-replicated 8-frame comparison, the advantage averaged one full point lower mean absolute error across all nine cross-seed pairings. A standard R3D-18 achieved the best accuracy in the study, with a mean absolute error of 3.99, while requiring 19% less CPU latency than the reference configuration. A 16-frame, period-4 R(2+1)D-18 cut reference cost in half with no statistically confirmed loss in accuracy. Removing temporal modeling entirely substantially reduced accuracy, with a mean absolute error of 5.65, setting a practical floor for how inexpensive this task can be. We release the code, cost-measurement protocol, and per-configuration results.
A. Pandey, K. Sharma, A. Shah· medRxiv· 0 citations
Background: Black individuals have a lower incidence of atrial fibrillation (AF) than White individuals despite a higher burden of many traditional cardiovascular risk factors. Differences in left atrial (LA) structure and function by race could partly explain the observed pattern of AF risk. Methods: This analysis included 4,576 (978 Black and 3,598 White) participants from the Atherosclerosis Risk in Communities (ARIC) study, followed between 2011 and 2021. The association of selected echocardiographic measures of LA structure and function with AF incidence was evaluated with race-specific Cox proportional hazards models with adjustment for sociodemographic and clinical covariates. Additional analyses assessed whether LA measures attenuated the association between race and incident AF. Results: The analysis included 778 AF cases (113 in Black and 665 in White participants, mean age 75 years). Larger LA size and worse LA function were associated with higher AF risk in both Black and White individuals, with most associations of similar magnitude in both groups, except for a slightly stronger association of LA reservoir strain in Black than White participants (Black: hazard ratio (HR) 0.89, 95% CI 0.86-0.92 per 1% increase; White: HR 0.94, 95% CI 0.92-0.95, p for interaction = 0.01). In the overall sample, White participants showed higher AF risk compared to Black participants (HR 1.59, 95% CI 1.24-2.03). Adjustment for most individual LA measures did not attenuate the association between race and AF risk. Conclusion: Larger LA size and worse LA function were associated with incident AF in both Black and White ARIC participants. However, these measures did not explain the lower AF incidence observed among Black participants. LA remodeling appears to be an important predictor of AF risk, but it is not the primary explanation for the Black-White AF paradox.
Yuchen Li, E. Soliman, Srishti Shrestha et al.· medRxiv· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
