LogoFDA 510(k) Search
Search Filters

Search Results

Found 1 results

510(k) Data Aggregation

    K Number
    K202847
    Device Name
    AccuBrain
    Manufacturer
    BrainNow Medical Technology Limited
    Date Cleared
    2021-01-15

    (112 days)

    Product Code
    LLZ
    Regulation Number
    892.2050
    Type
    Traditional
    Panel
    Radiology
    Reference & Predicate Devices
    K170981,K140828
    Predicate For
    N/A
    AI/MLSaMDIVD (In Vitro Diagnostic)TherapeuticDiagnosticis PCCP AuthorizedThirdpartyExpeditedreview
    Intended Use

    AccuBrain is a fully automated post-processing software that provides automatic labeling, visualization and volumetric quantification of hippocampus from a set of T1W MRIs and returns an analysis report.

    Device Description

    AccuBrain is a fully automated post-processing software that provides automatic labeling, visualization and volumetric quantification of hippocampus from a set of T1W MRIs and returns an analysis report. The resulting output is a morphometric report in PDF format. The software is suitable for use in both clinical trial research and routine patient care as a support tool for clinicians in assessment of structural MRIs.

    AccuBrain provides morphometric measurements based on 3D T1 MRI series. The output of the software includes volumes of hippocampus.

    The AccuBrain processing architecture includes an automated internal pipeline that performs artifact correction, segmentation, hippocampus quantification, volume calculation and report generation.

    Additionally, automated safety measures include automated quality control functions, such as DICOM check, age check and image resolution check and image quality check.

    AI/ML Overview

    Here's an analysis of the acceptance criteria and study detailed in the provided document for the AccuBrain device:

    Acceptance Criteria and Device Performance

    Acceptance CriteriaReported Device Performance
    Mean Dice coefficient for right hippocampus0.89 (std: 0.03)
    Mean Dice coefficient for left hippocampus0.89 (std: 0.03)
    Mean Dice coefficient for total hippocampus0.89 (std: 0.03)
    Mean intrascanner Coefficient of Variation (CV) for left hippocampus3.20%
    Mean intrascanner Coefficient of Variation (CV) for right hippocampus1.23%
    Mean percentage absolute volume differences (DIFF) for left hippocampus4.52%
    Mean percentage absolute volume differences (DIFF) for right hippocampus1.74%

    Study Details

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

    • Sample Size: 135 subjects.
    • Data Provenance: The data was provided by the EADC-ADNI HarP (European Alzheimer's Disease Consortium - Alzheimer's Disease Neuroimaging Initiative Harmonized Protocol for Hippocampal Segmentation). The document does not specify the country of origin, but ADNI is primarily a North American initiative with international collaborations, and EADC is European. The study states "The subjects upon whom the device was tested include healthy subjects, Alzheimer's disease patients and Mild Cognitive Impairment patients," implying retrospective use of pre-existing data.

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

    • The document states that the ground truth was "manual segmentation results." It does not specify the number of experts or their qualifications for the test set.

    4. Adjudication Method for the Test Set:

    • The document refers to "manual segmentation results" as the ground truth but does not describe an explicit adjudication method (e.g., 2+1, 3+1). The use of the term "manual segmentation results" in the singular might suggest a single expert, or a consensus that is not explicitly detailed.

    5. Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study:

    • A MRMC comparative effectiveness study was not explicitly described. The study focused on the standalone performance of AccuBrain against manual segmentation (ground truth) and compared its performance metrics (Dice coefficients, CV, DIFF) to previously established performance ranges or specific values of predicate devices. There is no mention of human readers improving with AI vs without AI assistance.

    6. Standalone (Algorithm Only Without Human-in-the-Loop Performance) Study:

    • Yes, a standalone study was done. The performance metrics (Dice coefficient, CV, DIFF) reported for AccuBrain are based on the algorithm's output compared to manual segmentation or repeated scans, without human intervention in the segmentation process.

    7. Type of Ground Truth Used:

    • Expert Consensus / Expert-driven segmentation: The ground truth for the test set was "manual segmentation results," which implies segmentation performed by human experts. The reference to the EADC-ADNI HarP further indicates that these manual segmentations likely followed a standardized protocol for consistency and accuracy.

    8. Sample Size for the Training Set:

    • The training set, referred to as the "atlas pool," consisted of 300 brain MRIs.

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

    • The ground truth for the training set (atlas pool) was established through "prior encoded radiologist-specified anatomy information for hippocampus." This indicates that human experts (radiologists) manually segmented or annotated the hippocampi in the 300 ATLAS MRIs, which then served as the ground truth for training the atlas-based segmentation method.
    Ask a Question

    Ask a specific question about this device

    Page 1 of 1