Search Results

The AccuMeasure™ System is intended to be used as an accessory in conjunction with an endoscope to measure observable anatomy and pathology in the gastrointestinal tract. The AccuMeasure™ System provides no therapeutic or diagnostic function.

The device is used in conjunction with an endoscope to measure objects on the mucosal surface of the gastrointestinal (GI) tract. The device consists of two components (1) the measuring device and (2) a processing unit. The measuring device consists of a through the scope probe that is connected to a handheld laser source. The processing unit includes a video grabber and a touch screen PC. The medical-grade PC comes with dedicated software for conducting measurements during an endoscopic procedure.

Here's a breakdown of the acceptance criteria and the study information for the AccuMeasure™ System, based on the provided text:

1. Table of Acceptance Criteria and Reported Device Performance

This table focuses on the "Accuracy Validation" from the Performance Testing - Bench section, as it's the primary quantitative performance claim for the device's main function.

Note: The clinical study did not assess the accuracy of the device due to limitations in determining the true length of structures in vivo. Its primary endpoints were safety and usability.

Study Details

Here's a breakdown of the specific information requested:

2. Sample size used for the test set and the data provenance (e.g. country of origin of the data, retrospective or prospective)

• Test Set Sample Size (for accuracy): The accuracy validation was conducted via bench testing using "phantoms" (simulated round and flat polyps between 5-15 mm in diameter). The exact number of simulated polyps or measurements is not specified, but the text states "for all sizes."
• Test Set Sample Size (for clinical study, usability/safety): The clinical study involved [redacted] patients from a healthcare facility in Israel.
  • [redacted] patients were excluded for poor preparation.
  • 100 patients had no polyps.
  • [redacted] patients were excluded due to an endoscope issue.
  • The remaining [redacted] patients had one or more polyps.
  • A total of [redacted] polyps were attempted to be measured.
  • [redacted] polyps could not be measured.
  • Of the remaining, physicians measured [redacted] polyps.
• Data Provenance:
  • Accuracy Test (Bench): Not explicitly stated, but implies internal testing by the manufacturer.
  • Clinical Study (Usability/Safety): Prospective; conducted at a healthcare facility in Israel.

3. Number of experts used to establish the ground truth for the test set and the qualifications of those experts (e.g. radiologist with 10 years of experience)

• Accuracy Test (Bench): For the bench testing, ground truth was the known, precise dimensions of the simulated polyps in the phantoms. No human experts were required to establish this ground truth, as it was inherent to the phantom design.
• Clinical Study (Usability/Safety): The clinical study focused on physician subjective assessments of usability and safety, not on establishing a ground truth for measurement accuracy. The measurements obtained in the clinical study were not used to assess the accuracy of the device against a "true" measurement. [redacted] physicians participated in the study; their qualifications are not specified beyond being "physicians."

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

• Accuracy Test (Bench): Not applicable. The ground truth was objective (known dimensions of phantoms).
• Clinical Study (Usability/Safety): The clinical study involved subjective user assessments. No adjudication method for ground truth determination was applicable, as ground truth for accuracy was not assessed in vivo. For adverse events, standard clinical monitoring was performed, but no formal adjudication of disagreements is described.

5. 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 comparative effectiveness study was done. The clinical study was solely for safety and usability, not comparative effectiveness with human readers or other measurement methods.
• The text does state that the device "demonstrated greater accuracy than other methods of assessing length during endoscopy procedures (e.g., visual estimation, biopsy forceps, and snares)" based on bench testing compared to previously published estimates of error ranges for visual estimation and biopsy forceps. This is not an MRMC study with human readers using the AI vs. without AI assistance.

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

• Standalone Performance (Accuracy): Yes, the "Performance Testing - Bench" section describes testing where the device (AccuMeasure™ System software) acquired images and made measurements of simulated polyps in phantoms. This was evaluating the algorithm's performance in a controlled, non-human-in-the-loop setting for its core function of measurement accuracy.
• Standalone Performance (Clinical): No. While the algorithm performs the calculation, it is used "in conjunction with an endoscope" and requires physician interaction to acquire images, position the laser, and
  drag target markers. It is not intended to be a standalone diagnostic device.

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

• Accuracy Test (Bench): Objective, precisely known physical dimensions of simulated polyps within custom-built phantoms.
• Clinical Study (Usability/Safety): The "ground truth" for the clinical study was physician subjective assessments via questionnaires for usability and observation for adverse events for safety. There was no ground truth for measurement accuracy in this study.

8. The sample size for the training set

• The provided text does not mention a specific training set sample size. The software is pre-calibrated for each endoscope model, but details about dataset size or how frequently "training" occurs are not given. This product description focuses on verification and validation, not development.

9. How the ground truth for the training set was established

• Given that the text does not mention a specific training set or process (beyond "pre-calibrated for each endoscope model to find its camera parameters and distortion coefficients"), the method for establishing ground truth for any potential training data is not described. It's possible that proprietary internal calibration procedures provided the necessary reference data without a formal "ground truth establishment" process in the context of an AI/ML training paradigm.

Ask a specific question about this device