EPIQ Series Diagnostic Ultrasound System, Affiniti Series Diagnostic Ultrasound System

The Smart Doppler View ID is an AI-driven software feature for Philips EPIQ and Affiniti ultrasound systems designed to streamline adult cardiac TTE exams by automatically selecting the appropriate Doppler calculation package group from an acquired spectrum. Its methodology relies on a trained algorithm that analyzes Doppler waveforms to suggest one of six cardiac views, demonstrated to have 97.5% accuracy, thereby enhancing workflow efficiency while allowing for user override.

The Philips EPIQ and Affiniti Series Diagnostic Ultrasound Systems introduce the Smart Doppler View ID, a software feature enhancing workflow in adult cardiac transthoracic echocardiography (TTE). It automates the selection of Doppler measurement calculation package groups on the system's touch screen by employing an artificial intelligence-based algorithm to analyze the user-acquired Doppler spectrum, thereby suggesting the relevant package group and streamlining subsequent measurements, a task previously done manually. This feature supports six specific Doppler calculation package groups (Aortic Valve, Mitral Valve, Tricuspid Valve, Pulmonic Valve, Venous Flow, and TDI Velocity & Ratio) and is a software-only update compatible with existing Philips transducers and systems on software version 11.0 or higher. Performance evaluation via a study with 1100 TTE cardiac clips from 400 subjects showed the AI algorithm achieved 97.5% accuracy in identifying the correct calculation package group compared to ground truth established by experienced clinicians, while still allowing users to manually override suggestions.

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 Smart Doppler View ID algorithm differentiate between complex or atypical Doppler spectra that might correspond to multiple valvular pathologies or unusual flow patterns?; What is the learning curve for sonographers to effectively integrate the Smart Doppler View ID into their existing workflow, and what is the typical rate of manual override observed in real-world clinical settings post-implementation?; Are there specific patient populations, cardiac conditions, or image qualities where the algorithm's 97.5% accuracy might be lower, and what mechanisms are in place to alert users in such scenarios?; Beyond the six currently supported cardiac calculation package groups, are there plans to expand the Smart Doppler View ID's capabilities to other Doppler applications or anatomical regions within echocardiography?; How does the Smart Doppler View ID algorithm handle variations in Doppler signal quality due to operator skill or patient factors, and does it provide feedback on spectrum quality for optimal AI processing?