Search Results

The EndoSound Vision System (EVS), when affixed to an endoscope, is intended to provide ultrasonic visualization of, and ultrasound guided therapeutic access to the Upper Gastrointestinal Tract including but not restricted to the organs, tissues, and subsystems: Esophagus, Stomach, Duodenum and underlying areas. The EVS, mounted on an endoscope, is introduced orally when indications consistent with the requirement for a GI procedure are met.

The EVS is a prescription only device to be used by a qualified physician.

The clinical environments where the system can be used include clinics, hospitals, and ambulatory surgery centers.

The EndoSound Vision System® (EVS) is an add-on ultrasound system designed to attach externally to an upper gastrointestinal (GI) endoscope (gastroscope/upper (EGD) endoscope). Once attached, it temporarily converts an EGD endoscope to a fully capable EUS endoscope. The EVS consists of:

• an electronic ultrasound beamformer/image analyzer (EVSScanner),
• a reusable transducer assembly (Ultrasound Transducer Module (UTM)), .
• . a cable/connector to interface the transducer to the beamformer (Transducer Extension Cable (TEC)).
• a sterile, disposable mounting kit with an operator control mechanism (Ultrasound Disposable Kit (UDK-T)) used to externally affix the EVS to a standard gastroscope/upper (EGD) endoscope and to provide needle and transducer angle control.

When attached to an upper GI endoscope, the EVS enables real-time ultrasound imaging, ultrasound quided needle aspiration and other EUS quided procedures within the upper gastrointestinal tract and surrounding organs. Native functions of the gastroscope including video imaging, articulation insufflation and suctioning are available after attachment of the EVS device. The system works in tandem with a PC computer/monitor/tablet connected via a USB 3.0 cable that runs custom software (EVSViewer) and provides the user interface and ultrasound display.

The provided document is a 510(k) summary for the EndoSound Vision System (EVS). It describes the device, its intended use, comparison to predicate devices, and various safety and performance tests. However, it does not contain details about an AI/algorithm performance study with specific acceptance criteria. The "study that proves the device meets the acceptance criteria" refers to a general series of verification and validation tests, and an animal study, primarily focused on the physical and functional aspects of the ultrasound system itself and not explicitly on the performance of a software algorithm that performs diagnostic analysis or image interpretation to address specific clinical outcomes (e.g., detecting pathologies).

The EVSViewer software mentioned primarily handles image processing and display, and integration with the EVS UTM probe, rather than performing AI-driven diagnostic interpretations. The document states: "The EVSViewer is the software run on a PC/monitor which processes the ultrasound data received from the EVSScanner module and displays the ultrasound image. It also functions to produce the user interface on the computer monitor." It further details minor changes to the software related to probe integration, recognition, usage counter, field of view lock, and splash screen/icons, none of which suggest an AI component for image analysis or diagnostic support.

Therefore, for the specific questions regarding acceptance criteria and a study proving an AI/algorithm-based device meets these criteria, the provided text does not offer the necessary information. It details traditional medical device testing for an ultrasound system.

However, I can extract information about the general safety and effectiveness studies conducted for the device as a whole, based on the provided text.

Here's a summary of the device's acceptance criteria and the studies mentioned, as much as can be gleaned from the text:

Device: EndoSound Vision System (EVS) - an add-on ultrasound system for upper GI endoscopes.

Acceptance Criteria and Reported Device Performance (General Device Performance, not AI-specific):

The document broadly refers to "Verification and validation testing" and "Successful results of testing" but does not explicitly list quantitative acceptance criteria in a table format for specific performance metrics of an AI. Instead, it discusses the device's functional equivalence and safety.

1. Sample sizes used for the test set and the data provenance:

• Test Sets:
  • Biocompatibility: Not specified beyond "components of the EVS in contact with the patient."
  • Reprocessing/Sterilization: Not specified (e.g., number of units tested).
  • Animal Testing: 3 live animal models. The regimen was implemented four times in each animal. Data provenance is not specified beyond "GLP lab study." It is a prospective study.
  • Phantom and In Vivo Study (for Ultrasound Scanning Comparison): Not explicitly quantified, but mentioned that "Phantom and in vivo study comparison images show similar scanning resolution, image quality and ultrasound image performance."
  • Verification and Validation Testing (general): Not specified.

2. Number of experts used to establish the ground truth for the test set and the qualifications of those experts:

• Animal Testing: "An expert endoscopist as the user under the supervision of the GLP quality director and the study director." The specific number and exact qualifications beyond "expert" are not detailed.
• For other tests (biocompatibility, reprocessing, etc.), expert interpretation is inherent in GLP lab studies, but specific numbers and qualifications of human experts establishing "ground truth" (in the sense of diagnostic image interpretation) are not mentioned.

3. Adjudication method (e.g. 2+1, 3+1, none) for the test set:

• Not explicitly described for any of the studies mentioned. The animal study mentions "an expert endoscopist as the user under the supervision of the GLP quality director and the study director," implying some form of oversight, but not a specific adjudication protocol for diagnostic ground truth.

4. If a multi-reader multi-case (MRMC) comparative effectiveness study was done, If so, what was the effect size of how much human readers improve with AI vs without AI assistance:

• No MRMC study was described. The document does not mention human reader studies, nor does it describe AI assistance for diagnostic tasks. The focus is on the substantial equivalence of the ultrasound imaging system and its physical components.

5. If a standalone (i.e. algorithm only without human-in-the-loop performance) was done:

• No standalone algorithm performance study was described. The "EVSViewer software" is described as processing and displaying ultrasound data, not performing automated diagnostic or interpretive functions.

6. The type of ground truth used (expert consensus, pathology, outcomes data, etc.):

• Animal study: "Clinically relevant damage was identified, either radiologically, endoscopically, or macroscopically" suggesting a combination of imaging and direct visual/physical assessment as ground truth for safety. For performance, it was a comparative assessment of image quality and device behavior relative to a predicate device.
• Other tests relying on established and regulated testing methodologies (e.g., ISO 10993 for biocompatibility, IEC 60601 for electrical safety).

7. The sample size for the training set:

• Not applicable. The document does not describe the development or training of an AI algorithm, so there is no training set mentioned. The "firmware modification" and "EVSViewer software" changes are related to device control and display, not AI.

8. How the ground truth for the training set was established:

• Not applicable. As no AI training set is mentioned, no ground truth establishment process for it is described.

In summary: The provided 510(k) summary focuses on the substantial equivalence of the EndoSound Vision System as an ultrasound imaging device to existing predicate devices, and its safety. It does not contain information about an AI/algorithm-based diagnostic device, nor does it specify AI-related acceptance criteria or studies to prove such criteria. The reported studies are typical for a hardware and basic software medical device submission.

Ask a specific question about this device