Philips Ultrasound, LLC · Class II · Cleared Jun 18, 2025
| K-number | K250886 |
| Device name | EPIQ Series Diagnostic Ultrasound Systems; Affiniti Series Diagnostic Ultrasound Systems |
| Applicant | Philips Ultrasound, LLC |
| Product code | IYN |
| Device class | Class II |
| Decision date | Jun 18, 2025 |
| Decision | Substantially Equivalent |
| Regulation | 892.1550 |
The EPIQ and Affiniti Series Diagnostic Ultrasound Systems with Smart View Select (SVS) v2 software are ultrasound diagnostic imaging devices that perform fluid flow analysis of the human body. SVS v2 is an automated software feature that uses artificial intelligence to classify and select optimal images for left ventricular analysis in adult cardiac transthoracic examinations, assisting users in image selection without manual review.
The subject devices use a deep learning neural network for view identification and optimal triplet (A4C, A2C, A3C) selection, with a revised heuristic logic step that preferentially selects triplets of shallower scan depths (12–20 cm restriction for proposed device vs. predicate's 10 cm minimum with 1 cm variation constraint). The classification engine is AI-based while the selection algorithm is non-AI. No hardware changes were made; existing cleared Philips transducers are used.
IEC 62304 (Medical device software – Software life cycle processes, 2006 + A 2015), IEC 62366-1 (Usability engineering to medical devices, Edition 1.1 2020-06), ISO 14971 (Risk management for medical devices, 2019), and FDA Guidance for Industry and FDA Staff – Marketing Clearance of Diagnostic Ultrasound Systems and Transducers (February 2023).
The proposed devices are substantially equivalent because: (1) indications for use, user environment, and regulatory classification are identical to predicate K240850; (2) performance testing via retrospective analysis demonstrated strong correlation between SVS-selected clips and manually selected clips for ejection fraction (r=0.891, 95% CI 0.851–0.920) and global longitudinal strain (r=0.906, 95% CI 0.871–0.931), both exceeding the predefined acceptance criterion of >0.8; and (3) the revision to the depth-selection heuristic does not raise new safety or effectiveness questions, as the algorithm remains functionally similar and produces clinically equivalent results.
View the full FDA submission: accessdata.fda.gov