FDA 510(k) Search - By Innolitics (SaMD and AI Experts)

The 4Sight is a multi-purpose diagnostic ophthalmic ultrasound system intended to make axial length, anterior chamber depth, lens thickness and corneal thickness measurements of the internal structure of the eye.

Diagnostic ultrasound imaging or fluid flow analysis of the human body as follows: Ophthalmic

Device Description

The 4sight device is designed to combine the functionalities of Accutome's A-scan Plus (K123349), B-scan Plus (K070943) and Accupach Pachymeter (K042752) devices into a combined, stand-alone platform. The device performs axial length measurements, corneal thickness measurements and images the internal structure of the eye using these 3 modalities. The 4Sight is a stand-alone system which runs on a Windows 8 platform consisting of a 4Sight control unit, transducers, footswitch, and keyboard.

AI/ML Overview

Here's an analysis of the acceptance criteria and study information provided in the document for the Accutome 4Sight device:

1. Table of Acceptance Criteria and Reported Device Performance

Test/Criteria	Acceptance Criteria	Reported Device Performance
A-scan Physical Accuracy Test	Well within 0.1mm	Both the A-scan function on the 4Sight and the predicate device were well within the 0.1mm acceptance criteria.
Pachymetry Comparison Test	All data met established acceptance criteria	All of the data met the established acceptance criteria.
Software Validation	317 software requirements met	All 317 software requirements were tested and validated. All bugs, except one minor anomaly, were corrected.
External Validation (Safety)	Ratings of 4 or 5 (5=best) on a 1-5 scale for overall rating, no 1s or 2s. Safety ranked 5.	Safety was ranked 5 by every user. The "Overall Rating" of the device was 4 or 5 by all users. There were no ratings of 1 or 2 in any of the tests.
External Validation (Usability/Functionality)	Ratings of 4 or 5 (5=best) on a 1-5 scale for overall rating, no 1s or 2s. Usability judged equivalent to predicate.	The "Overall Rating" of the device was 4 or 5 by all users. There were no ratings of 1 or 2 in any of the tests. Usability was judged to be equivalent to the predicate device during internal validation.

2. Sample Size Used for the Test Set and Data Provenance

A-scan Physical Accuracy Test: The document does not specify a numerical sample size for this test. It refers to "the A-Scan function on the 4Sight" and "the predicate device," suggesting comparative measurements, but not how many such measurements were taken.
Pachymetry Comparison Test: The document does not specify a numerical sample size for this test. It refers to "critical parameter results between the two devices," implying multiple comparisons, but the exact number is not provided.
External Validation Tests: The device was evaluated at 4 different locations. The document does not specify the number of individual cases or patients involved.
Data Provenance: Not explicitly stated regarding country of origin. The study appears to be prospective for the validation testing, as it involved evaluating the new 4Sight device.

3. Number of Experts Used to Establish the Ground Truth for the Test Set and Their Qualifications

A-scan Physical Accuracy Test: Ground truth was established by "physical distance measured by a mechanical dial indicator." This does not involve human experts in setting the ground truth, but rather a direct physical measurement.
Pachymetry Comparison Test: Ground truth for this test involved comparing the 4Sight's results to those of the predicate device (Accupach VI) using the "same input test fixture and values." The predicate device likely serves as the reference, which itself would have been validated against a form of ground truth or clinical standard.
External Validation Tests: "Experienced technicians/ophthalmologists" evaluated the device at 4 different locations. The exact number of experts is not specified beyond "users." Their specific years of experience or training are also not detailed, only that they were "experienced."

4. Adjudication Method for the Test Set

No explicit adjudication method (e.g., 2+1, 3+1) is mentioned for any of the tests.
For the External Validation Tests, the qualitative feedback (ratings) from "every user" or "all users" was aggregated to determine if the acceptance criteria were met. This implies a consensus or majority opinion approach rather than a formal adjudication process for discordant interpretations.

5. If a Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study was Done

No, a formal MRMC comparative effectiveness study, as typically understood for AI evaluation, was not conducted.
The document describes a "Pachymetry Comparison Test" and an "A-scan Physical Accuracy Test" that compare the 4Sight to its predicate devices, but this is a device-to-device comparison, not an AI-assisted vs. unassisted human reader comparison.
The External Validation Tests involved multiple users, but the goal was to assess device usability and safety, not to quantify the effect size of AI on reader performance. There is no mention of "AI" in this context, as the device is an ultrasound system making measurements.

6. If a Standalone (i.e. algorithm only without human-in-the-loop performance) was Done

Yes, the performance of the device's measurement algorithms appears to have been evaluated in a standalone manner.
The A-scan Physical Accuracy Test directly measures the accuracy of the A-scan function against a physical standard.
The Pachymetry Comparison Test compares the 4Sight's output with that of a predicate device under controlled input conditions, essentially evaluating the algorithm's output.
The device itself is described as a "stand-alone system."

7. The Type of Ground Truth Used

A-scan Physical Accuracy Test: Direct physical measurement (physical distance measured by a mechanical dial indicator).
Pachymetry Comparison Test: Comparison against a predicate device's measurements using the same input. The predicate device's measurements are implicitly treated as the ground truth for this comparison.
Software Validation: Ground truth was meeting the 317 specified software requirements.
External Validation Tests: Expert opinion/feedback (ratings from experienced technicians/ophthalmologists) regarding usability and safety.

8. The Sample Size for the Training Set

The document does not describe the development of an AI algorithm that would require a "training set." The 4Sight device is described as combining functionalities and algorithms from existing predicate devices (A-scan Plus, B-scan Plus, Accupach Pachymeter).
The software modifications were primarily for integrating these functionalities into a single standalone unit and graphical interface, not for developing new measurement algorithms that would typically rely on training data. Therefore, the concept of a training set as applied to AI/ML is not relevant here.

9. How the Ground Truth for the Training Set Was Established

As the device does not appear to employ new AI/ML algorithms requiring a training set in the conventional sense, this information is not applicable and therefore not provided in the document. The algorithms for critical functions were stated to be "exactly the same as in the predicate devices."

Summary

Regulation Number and Section

§ 892.1560 Ultrasonic pulsed echo imaging system.

(a)
Identification. An ultrasonic pulsed echo imaging system is a device intended to project a pulsed sound beam into body tissue to determine the depth or location of the tissue interfaces and to measure the duration of an acoustic pulse from the transmitter to the tissue interface and back to the receiver. This generic type of device may include signal analysis and display equipment, patient and equipment supports, component parts, and accessories.(b)
Classification. Class II (special controls). A biopsy needle guide kit intended for use with an ultrasonic pulsed echo imaging system only is exempt from the premarket notification procedures in subpart E of part 807 of this chapter subject to the limitations in § 892.9.