LVivo Software Application

The LVivo Software Application is designed to provide automated, non-invasive analysis of ultrasound images for quantitative assessment of cardiac structures and function, including left and right ventricular parameters and bladder volume. Its methodology leverages DICOM data and incorporates a neural network-enhanced algorithm for ejection fraction calculation, aiming to deliver reliable medical parameters for patients with suspected disease.

The LVivo Software Application is a medical software platform designed for the non-invasive processing and automated analysis of ultrasound images, particularly echocardiograms. It utilizes DICOM movie files to perform global and segmental evaluations of cardiac structures, primarily the left ventricle (LV) and right ventricle (RV), and also includes a module for bladder volume measurement. The system's methodology involves automated analysis to calculate key medical parameters such as ejection fraction (EF), segmental wall motion (SWM), and global longitudinal strain (GLS) from various echocardiographic views including apical (2CH, 4CH, 3CH) and parasternal short axis (SAX). A significant aspect of its science involves a modified algorithm in its LVivo EF module, which incorporates neural network technology to enhance its analytical capabilities. Performance evaluation of the LVivo Software Application was conducted using a dataset from 100 adult patients undergoing routine transthoracic echocardiography, encompassing individuals with normal left ventricular function and those with coronary artery disease. The software demonstrated its ability to meet predefined acceptance criteria for ejection fraction analysis (e.g., biplane EF correlation ≥80%) and for segmental wall motion and global longitudinal strain analysis (e.g., sensitivity ≥75% compared to reference methods). These results indicate that the software, including its neural network-enhanced algorithm, effectively provides automated measurements and functional assessments of cardiac parameters and bladder volume, supporting its intended use in patients with suspected disease.

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 neural network-enhanced LVivo EF algorithm compare specifically to the previous algorithm version used in the predicate device, particularly in challenging or borderline patient cases?; What is the inter-observer and intra-observer variability of measurements obtained using the LVivo Software Application for all its modules (LV EF, SG, SAX, RV, Bladder) when compared to manual expert analysis?; How does the LVivo software handle ultrasound images of varying quality, and what are its performance metrics, such as failure rates or accuracy degradation, with suboptimal image inputs?; Beyond coronary artery disease and normal LV function, has the LVivo software been validated in other specific cardiac pathologies (e.g., various cardiomyopathies, valvular heart disease) or diverse patient populations, such as pediatric patients?; What are the specific implications of the 'Android option for LVivo EF' on clinical workflow, data security, and the range of ultrasound devices it can effectively interface with in a practical setting?