Myocardial Stiffness Tracking & Assessment Toolkit Driven by Artificial Intelligence For Reproducible Benchmarking of Temporal Segmentation and Shear-Wave Velocity Stabilization on Synthetic Data

The article presents MyoSTAT.AI, a new artificial intelligence-driven tool designed to improve echocardiographic assessment. The system automates cardiac segmentation and shear-wave velocity estimation in real-time, reducing operator variability and processing time. The research team tested four U-Net-based architectures on synthetic data containing 1,200 frames for segmentation and 900 velocity-field cases. The ConvLSTM variant achieved the best results with Dice accuracy of 0.994 and IoU of 0.987, representing an 18.6% improvement over the baseline 2D model. The 2.5D model with three frames achieved Dice 0.983 with significantly lower latency. Deployment on RTX 3060 hardware achieved 341 FPS and on Jetson Orin Nano 90.4 FPS. The proposed stabilization method reduced shear-wave velocity variability by 63.8% without loss of edge preservation. However, the authors emphasize this is a proof-of-concept on synthetic data, and validation on real echocardiographic data is required before clinical use.