(70 days)
The Philips Lumify Diagnostic Ultrasound System is intended for diagnostic ultrasound imaging in B (2D), Color Doppler, Combined (B+Color), Pulsed Wave Doppler (PWD), and M-modes.
It is indicated for diagnostic ultrasound imaging and fluid flow analysis in the following applications: Fetal/Obstetric, Abdominal, Pediatric, Cephalic, Urology, Gynecological, Cardiac Fetal Echo, Small Organ, Musculoskeletal, Peripheral Vessel, Carotid, Cardiac, Lung,
The Lumify system is a transportable ultrasound system intended for use in environments where healthcare is provided by healthcare professionals.
The Philips Lumify Diagnostic Ultrasound System (Lumify) is a mobile, durable, and reusable, software-controlled medical device, which is intended to acquire high-resolution ultrasound data and to display the data in B (2D), Pulsed Wave Doppler, Color Doppler, Combined (B+ Color), and M modes.
The Lumify Diagnostic Ultrasound System (Android) utilizes:
- A commercial off-the-shelf (COTS) Android mobile device (smart phone or tablet)
- The Philips Ultrasound Lumify software running as an application on the COTS device
- The Philips C5-2 Curved array USB transducer
- The Philips L12-4 Linear array USB transducer
- The Philips S4-1 Sector array USB transducer
- Lumify Micro B Transducer Cable
- Lumify Micro C Transducer Cable
The Lumify system is compatible with iOS or Android operating systems. The Lumify system software provides various imaging features, including an Android-specific feature with artificial intelligence (AI) based, Auto EF Quantification (ejection fraction) technology during cardiac imaging.
Here's a breakdown of the acceptance criteria and the study details for the Philips Lumify Diagnostic Ultrasound System with Auto EF Quantification, based on the provided FDA 510(k) summary:
1. Table of Acceptance Criteria and Reported Device Performance
| Acceptance Criteria | Reported Device Performance |
|---|---|
| Correlation of LVivo EF Ejection Fraction (EF) measurements with average manual tracing results. | Strong correlation demonstrated: r = 0.82, 95% CI (0.72, 0.88). The endpoint criteria were met. |
| Correlation of LVivo EF End-Diastolic Volume (EDV) measurements with average manual tracing results. | Strong correlation demonstrated: r = 0.95, 95% CI (0.91, 0.96). |
| Correlation of LVivo EF End-Systolic Volume (ESV) measurements with average manual tracing results. | Strong correlation demonstrated: r = 0.94, 95% CI (0.90, 0.96). |
| Percentage of clips successfully processed automatically by LVivo EF. | 76 out of 80 clips (95%) were automatically processed. |
2. Sample Size Used for the Test Set and Data Provenance
- Sample Size for Test Set: 80 patients' Apical 4 Chamber (A4CH) view clips.
- Data Provenance: The data were acquired with the Lumify Diagnostic Ultrasound System, specifically for this clinical performance study. Patients were selected based on eligibility, and data were acquired consecutively for patients with normal and impaired LV function. This suggests a prospective acquisition for the purpose of the study. The document does not specify the country of origin for the data.
3. Number of Experts Used to Establish the Ground Truth for the Test Set and Qualifications
The ground truth was established by "manual tracing performed by sonographers."
- Number of Experts: The document refers to "sonographers" in the plural, but does not specify the exact number of sonographers involved in the manual tracing.
- Qualifications of Experts: The document implies that these were qualified "sonographers" experienced in echocardiographic LV function evaluation, but does not provide specific qualifications (e.g., years of experience, board certification). It can be inferred that they are healthcare professionals who routinely perform this task.
4. Adjudication Method for the Test Set
The ground truth for EF, EDV, and ESV was established by the "average results by manual tracing." This implies that IF multiple sonographers performed the manual tracings, their results were averaged. However, it does not explicitly state an adjudication method like 2+1 or 3+1 (where discrepancies are resolved by a third expert or consensus). It refers to "the average results by manual tracing," suggesting a quantitative aggregation rather than a specific adjudicative consensus process if multiple readers were used. If only one sonographer performed the tracing for each case, no adjudication would be necessary.
5. Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study
There is no mention of a Multi-Reader Multi-Case (MRMC) comparative effectiveness study being done to evaluate how much human readers improve with AI vs. without AI assistance. The study focuses on the agreement between the AI's automated measurements and a "manual tracing" ground truth, not on reader performance with and without AI assistance.
6. Standalone (Algorithm Only) Performance
Yes, a standalone study was done. The clinical performance study directly compared the "automated EF evaluation by LVivo EF" (the algorithm's performance) against "Ejection Fraction (EF) evaluation by manual tracing performed by sonographers" (the ground truth). The results (correlation coefficients) reflect the algorithm's performance without a human-in-the-loop scenario. The LVivo EF automatically processed 95% of the clips, indicating its standalone capability.
7. Type of Ground Truth Used
The ground truth used was expert consensus/manual tracing. Specifically, it was defined as "Ejection Fraction (EF) evaluation by manual tracing performed by sonographers" and "the average results by manual tracing" for EF, EDV, and ESV. This is considered an expert-derived ground truth based on conventional, established methods for echocardiographic LV function evaluation.
8. Sample Size for the Training Set
The document explicitly states: "The data used for clinical performance study were completely distinct from that used during training of the algorithm, and there was no overlap between the two data sets." However, it does not provide the sample size for the training set used to develop the LVivo EF algorithm.
9. How the Ground Truth for the Training Set was Established
The document states that the clinical performance study data were "completely distinct from that used during training of the algorithm," but it does not describe how the ground truth for the training set was established. It implies that such a training process occurred ("training of the algorithm"), but details on its ground truth are not provided in this summary.
§ 892.1550 Ultrasonic pulsed doppler imaging system.
(a)
Identification. An ultrasonic pulsed doppler imaging system is a device that combines the features of continuous wave doppler-effect technology with pulsed-echo effect technology and is intended to determine stationary body tissue characteristics, such as depth or location of tissue interfaces or dynamic tissue characteristics such as velocity of blood or tissue motion. This generic type of device may include signal analysis and display equipment, patient and equipment supports, component parts, and accessories.(b)
Classification. Class II.