Myocardial Strain Software Application

The Myocardial Strain Software Application is designed for qualified medical professionals to perform qualitative and quantitative evaluation of cardiovascular MR images, providing measurements of 2D LV myocardial function to aid in diagnosing suspected heart disease. The software employs a feature tracking algorithm based on a 2D nearly incompressible deformable model, initiated by user-defined myocardial contours, to compute parameters like strain, displacement, and velocity from SSFP cine CMR images.

The Myocardial Strain Software Application, also known as the Strain Module, is a medical image processing system developed by Circle Cardiovascular Imaging. It is intended for the qualitative and quantitative analysis of cardiovascular magnetic resonance (CMR) images, specifically those acquired using SSFP cine imaging. The software assists qualified medical professionals in evaluating 2D left ventricular (LV) myocardial function by providing a suite of measurements such as displacement, velocity, strain (including Global Longitudinal Strain and Global Circumferential Strain), strain rate, time to peak, and torsion. These measurements are used to obtain diagnostic information for patients with suspected heart disease, serving as an aid to the existing standard of care and not replacing other diagnostic tools. The results can be visualized on-screen, saved for later review, and the module can be integrated into broader cardiac image viewing platforms. The core methodology of the Strain Module relies on CMR feature tracking. Users initiate the process by confirming endocardial and epicardial contours in a reference phase of the cardiac cycle. The software then utilizes a 2D nearly incompressible deformable model algorithm to track these features throughout the cardiac cycle, with deformation determined by control points on the myocardial wall that are detected based on feature-tracked boundaries and an incompressibility constraint. This process is purely mathematical and does not involve artificial intelligence or machine learning. The device's performance and output validation were conducted using a combination of analytical phantoms, realistic phantoms with known deformations, real MRI data, and comparison against manual tracking by expert readers, ensuring compliance with relevant ISO standards and FDA guidance for quantitative imaging.

AI Regulatory Concerns

Powered by

Jev reads the full FDA letter and returns calibrated concern probabilities per regulatory category. These are AI estimates, not regulatory advice.

• Q1. How does the performance of the 2D nearly incompressible deformable model compare to other strain analysis algorithms (e.g., speckle tracking echocardiography or other CMR-based techniques) in terms of accuracy and reproducibility across different patient populations and image qualities?; What is the sensitivity and specificity of the derived strain parameters (GLS, GCS) using this software for detecting specific cardiac pathologies, and are there established cut-off values for these parameters based on studies using this specific module?; The validation involved phantoms and real MRI data; could you elaborate on the characteristics of the 'real MRI data' patient cohort (e.g., healthy volunteers, specific pathologies) and the inter-observer variability of contour placement affecting the strain results?; While the software is vendor-independent for CMR acquisition, are there any known impacts of varying acquisition parameters (e.g., temporal resolution, spatial resolution, slice thickness) on the accuracy of the strain measurements?; Given the module's reliance on user-defined contours in a reference phase, what is the learning curve for new users to achieve consistent and accurate contouring, and does the software offer any tools to guide or validate user inputs?