EPIQ Series Diagnostic Ultrasound System; Affiniti Series Diagnostic Ultrasound System

SVS v2 Contrast is an automated software feature added to Philips EPIQ and Affiniti Series Diagnostic Ultrasound Systems. It assists users in automatically selecting optimal cardiac ultrasound images for left ventricle analysis when contrast agent is used in adult transthoracic echocardiography exams, then launches existing analysis applications (2D Auto EF Advanced, AutoStrain LV, or 2D Auto LV) without requiring manual view selection.

SVS v2 Contrast revises the predicate SVS v1 algorithm to include optimization for contrast-enhanced imaging. The classification engine uses the same Deep Learning LVivo Seamless AI neural network as the predicate; the selection algorithm is non-AI and considers view classification and image depth. The software runs on existing EPIQ and Affiniti hardware with existing cleared transducers and prioritizes contrast image pairs but can fall back to non-contrast images if necessary.

IEC 62304 (Medical device software – Software life cycle processes), IEC 62366-1 (Medical devices – Application of usability engineering), ISO 14971 (Medical devices – Application of risk management). Testing followed FDA Guidance for Industry and FDA Staff – Marketing Clearance of Diagnostic Ultrasound Systems and Transducers (February 2023).

The proposed device uses the same LVivo Seamless AI neural network and operates within the same predicate system architecture (EPIQ/Affiniti with existing transducers). The primary algorithmic change—extending clip selection to contrast imaging—represents a narrow refinement rather than a fundamental change in function or risk profile. Performance testing demonstrated correlation of 0.953 (95% CI 0.917–0.974) for biplane ejection fraction measurement against manually selected ground truth, exceeding the acceptance criterion of >0.8. The feature maintains user override capability and does not introduce new hardware, new anatomical indications, or safety concerns beyond those addressed in the predicate submissions.