Search Results

AZE Phoenix software is intended as a radiological review station and for comparison of medical images from multiple imaging modalities and/or time-points. When interpreted by a trained radiologist the images displayed in Phoenix may be used as a basis for diagnosis. AZE Phoenix supports review and comparison by collecting all image data associated with a patient from disparate data servers, modalities, and time-points in a patient centric data list. Users select images from the patient data list for qualitative visual inspection of individual study images and comparison of aligned side-by-side studies or fused visualizations. Users can also make quantitative anatomical measurements in the images including RECIST guideline measurements and save measurements to track changes over time in graphical plots. After selection and comparison of time-series images, individual images may be loaded into AZE VirtualPlaceTM workstation which is launched directly from Phoenix for series specific analysis. In summary, Phoenix enables medical image inspection and diagnosis along with qualitative and quantitative comparison of patient images over time followed by in depth analysis in AZE VirtualPlaceTM workstation.

AZE Phoenix is not indicated for mammography use. Lossy compressed mammographic images and digitized film screen images must not be reviewed for primary image interpretations. Mammographic images may only be interpreted using an FDA-approved monitor that offers at least 5 Mpixel resolution and meets other technical specifications reviewed and accepted by FDA.

AZE Phoenix is a distributed client-server radiological review station that allows easy selection, review, processing and time-series comparison of multi-modality DICOM-compliant radiological images of patients. The AZE Phoenix software accepts, transfers, displays, stores, and digitally processes DICOM medical images from a variety of diagnostic imaging systems (such as CT, MRI, or image archives) for viewing image manipulation, communication, printing, and quantification. AZE Phoenix queries, retrieves, and sends DICOM compliant radiological images of patients to and from any number of DICOM Picture Archiving and Communication Systems (PACS) images sources. Phoenix collects all retrieved image series from all time-points and imaging modalities associated with a patient in a single patient centric list. Phoenix capabilities include: side-by-side and fusion image display; two-dimensional length, size, and angle measurements; Response Evaluation Criteria in Solid Tumors (RECIST) guideline measurements; image alignment by registration; tagging images to rapidly query and display image; customizable hanging protocol display layouts; and the ability to push specific image series to AZE VirtualPlace™ workstation (cleared under K060453) for in depth series specific analysis.

The AZE Phoenix device is a radiological review station for the comparison of medical images. The submission provided does not contain explicit acceptance criteria or a detailed study report with specific performance metrics against those criteria. Instead, it demonstrates substantial equivalence to predicate devices primarily through a comparison of technological characteristics and a general statement of measurement validation.

Here's an analysis of the provided information, framed to answer the requested points:

AZE Phoenix Device: Analysis of Acceptance Criteria and Performance

1. Table of Acceptance Criteria and Reported Device Performance

As no explicit acceptance criteria are defined in terms of quantitative performance metrics, the "acceptance criteria" can be inferred from the validation tests described and the "reported device performance" is the statement of no significant differences or improvement observed.

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

• Sample Size: Not explicitly stated. The submission mentions "CD ROM phantom CT images, clinical CT and MRI images" for measurement validation and "human breath cycle CT time series images" for alignment validation. No specific number of cases or images is provided for these test sets.
• Data Provenance: Not explicitly stated. The data is described as "CD ROM phantom CT images" and "clinical CT and MRI images" which are generic descriptions. No information on the country of origin of the data or whether it was retrospective or prospective is provided.

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

Not explicitly stated. The submission does not detail the process of establishing ground truth for the validation tests or the number and qualifications of experts involved.

4. Adjudication Method for the Test Set

Not explicitly stated. Given the comparison of measurements against a predicate device and registration methods, it's implied that the "ground truth" for comparison was either derived from the predicate device's output or established by a defined reference method, but no specific adjudication process (e.g., 2+1, 3+1) is mentioned.

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

No, a multi-reader multi-case (MRMC) comparative effectiveness study was not explicitly described or mentioned in the provided text. The performance testing focuses on technical validation and comparison to a predicate device's measurements, not on the improvement of human readers with AI assistance.

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

Yes, a standalone performance study was conducted. The "Performance Testing (Bench)" describes validation of the device's measurement capabilities ("geometric measurements, intensity measurements and free drawing RECIST measurements") and image alignment independent of human interpretation as the primary outcome. These tests compare the device's output directly to known values (phantoms) or the output of a predicate device.

7. The Type of Ground Truth Used

The types of ground truth used can be inferred as:

• Reference Measurements/Calculations: For "geometric measurements, intensity measurements and free drawing RECIST measurements," the ground truth likely came from:
  • Phantom data: For the "CD ROM phantom CT images," the ground truth would be the known, precisely defined dimensions and properties within the phantom.
  • Predicate device output: For "clinical CT and MRI images comparing subject and predicate (VirtualPlace) geometric measurements, intensity measurements and free drawing RECIST measurements," the predicate device's measurements likely served as the reference for comparison, indicating "no significant differences" if the results were within acceptable variability.
• Reference Registration Methods: For "Alignment and fusion image display validation using human breath cycle CT time series images," the "full registration to liver target registration" suggests a specialized, likely more accurate, registration method or expert-derived target registration as the reference ground truth for evaluating the device's alignment improvement.

8. The Sample Size for the Training Set

Not applicable. The AZE Phoenix device described here is a radiological review station with image processing capabilities, not an AI/ML-based algorithm that typically requires a distinct training set. The submission does not mention any machine learning components that would necessitate a training set. The "validation" performed is for the accuracy and functionality of its measurement and display tools.

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

Not applicable, as no training set for an AI/ML algorithm is mentioned or implied.

Ask a specific question about this device