LogoFDA 510(k) Search
Search Filters

Search Results

Found 1 results

510(k) Data Aggregation

    K Number
    K213891
    Device Name
    RealNow
    Manufacturer
    YIWEI Medical Technology Co., Ltd
    Date Cleared
    2023-02-24

    (437 days)

    Product Code
    LLZ
    Regulation Number
    892.2050
    Type
    Traditional
    Panel
    Radiology
    Reference & Predicate Devices
    K121447
    Predicate For
    N/A
    AI/MLSaMDIVD (In Vitro Diagnostic)TherapeuticDiagnosticis PCCP AuthorizedThirdpartyExpeditedreview
    Intended Use

    RealNow is an image processing software package to be used by trained professionals including physicians. The software runs on a standard "off-the-shelt" computer or a virtual platform, such as VMware, and can be used to perform image viewing, processing, analysis and communication of computed tomography (CT) perfusion scans of the brain. Data and images are acquired through DICOM compliant imaging devices.

    RealNow provides viewing, analysis and communication capabilities for functional and dynamic imaging datasets that are acquired with CT Perfusion imaging protocols.

    Analysis includes calculation of parameters related to tissue flow (perfusion) and tissue blood volume. Results of image processing which include CT perfusion parameter maps generated in the standard DICOM format and may be viewed on existing radiological imaging viewers.

    Device Description

    RealNow image analysis includes calculation of the following perfusion related parameters:

    • Cerebral Blood Flow (CBF)
    • Cerebral Blood Volume (CBV)
    • Mean Transit Time (MTT)
    • Residue function time-to-peak (TMax)
    • Arterial Input Function (AIF)

    The primary users of RealNow are medical imaging professionals who analyze dynamic CT perfusion studies. The results of image analysis produced by RealNow should be viewed through appropriate diagnostic viewers when used in clinical decision making.

    AI/ML Overview

    The RealNow device is an image processing software package for CT perfusion scans of the brain. The acceptance criteria and the study proving the device meets these criteria are detailed below. The study's "ground truth" for non-clinical testing was the output of a legally marketed predicate device, RAPID (K121447).

    1. Table of Acceptance Criteria and Reported Device Performance

    The acceptance criteria for RealNow were established by comparing its output to the RAPID predicate device. The performance was assessed through pixel-by-pixel comparison and volumetric analysis.

    Acceptance Criterion (for each scan)Reported Device Performance
    Intraclass Correlation Coefficient (ICC) > 0.75 for all voxels within the brain tissue for parametric maps.CBF, CBV, MTT, Tmax Parameter Maps: All proportion of acceptable maps (ICC > 0.75) were > 90% (p=0.0009 for CBF and CBV, p=0.0196 for MTT and Tmax). Specific ICC values are not provided, only the proportion of maps meeting the threshold.
    Assessment of bias for volumetric comparison (e.g., CBF<20%, CBF<30%) between RealNow and RAPID.Rapid_Volume (CBF<20%) vs RealNow_Volume (CBF<20%) & Rapid_Volume (CBF<30%) vs RealNow_Volume (CBF<30%): Assessment revealed bias is not significant. This indicates good agreement in these volumetric measurements.
    Volumetric comparison for Tmax abnormality between RealNow and RAPID.Tmax Abnormality: Assessment revealed a negative bias (mean Rapid Volume being smaller than mean Tmax volumetry) that increased as the average volume increased. Despite this bias, the document concludes "Good correlation in behavior between Rapid and RealNow volumetry was found." This suggests that while there may be a systematic difference, the trend and overall behavior are consistent.
    Mismatch volume accuracy parameters for RealNow compared to RAPID.Mismatch volumes (total data): < 50 ml. Mismatch volumes (stroke data): < 25 ml. The document states: "For Proposed Device RealNow mismatch analysis, mismatch volume has been established within the following accuracy parameters: < 50 ml for mismatch volumes (total data); < 25 ml for mismatch volumes (stroke data)." The phrasing suggests these were the target acceptance criteria, and the overall conclusion of "Good agreement and correlation in behavior between Rapid and RealNow was found" implies these were met.

    2. Sample size used for the test set and the data provenance

    • Sample Size: Not explicitly stated as a number of unique patient cases. The study refers to "CTP images (Real-World Evidence)" which are "paired uploaded to RealNow and RAPID for perfusion analysis." The number of individual scans analyzed for comparison of parametric maps is implied to be enough to achieve the reported p-values for the proportion of acceptable maps.
    • Data Provenance: "Real-World Evidence" images. The country of origin is not specified, nor is whether the data was retrospective or prospective. Given the comparison to a predicate device's existing output, it is highly likely that the data was retrospective.

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

    No human experts were used to establish the ground truth for the test set. The ground truth was established by the output of the legally marketed predicate device, RAPID.

    4. Adjudication method for the test set

    No human adjudication method (e.g., 2+1, 3+1) was used for the test set. The comparison was directly between the RealNow algorithm output and the RAPID algorithm output.

    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 involving human readers was mentioned or performed in this submission. The study focuses solely on the non-clinical performance comparison between the RealNow software and the predicate RAPID software.

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

    Yes, a standalone (algorithm only) performance study was done. The entire non-clinical test described compares the RealNow algorithm's output directly against the RAPID algorithm's output (which serves as the "ground truth" reference).

    7. The type of ground truth used

    The ground truth used was the output of a legally marketed predicate device (RAPID). This is a form of "reference standard" or "comparator" ground truth, where the performance is benchmarked against an already accepted technology rather than an independent gold standard like pathology or long-term outcomes.

    8. The sample size for the training set

    The document does not provide any information regarding a training set or its sample size. This is a non-clinical 510(k) submission, and the focus is on the validation of the device's performance against a predicate rather than the specifics of its development or training data.

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

    Information on how the ground truth for any potential training set was established is not provided in this document.

    Ask a Question

    Ask a specific question about this device

    Page 1 of 1