Search Results

The uMR 680 system is indicated for use as a magnetic resonance diagnostic device (MRDD) that produces sagittal, transverse, coronal, and oblique cross sectional images, and that display internal anatomical structure and/or function of the head, body and extremities. These images and the physical parameters derived from the images when interpreted by a trained physician yield information that may assist the diagnosis. Contrast agents may be used depending on the region of interest of the scan.

The uMR 680 is a 1.5T superconducting magnetic resonance diagnostic device with a 70cm size patient bore. It consists of components such as magnet, RF power amplifier, RF coils, gradient power amplifier, gradient coils, patient table, spectrometer, computer, equipment cabinets, power distribution system, internal communication system, and vital signal module etc. The uMR 680 Magnetic Resonance Diagnostic Device is designed to conform to NEMA and DICOM standards.

Here's a breakdown of the acceptance criteria and study details for the DeepRecon algorithm found in the provided FDA 510(k) summary:

1. Table of Acceptance Criteria and Reported Device Performance (DeepRecon)

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

• Sample Size for Test Set: 68 US subjects.
• Data Provenance: The test data was collected from various clinical sites in the US during separate time periods and on subjects different from the training data. The data specifically indicates demographic distributions for US subjects across various genders, age groups, ethnicities, and BMI groups.

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

The document mentions that "DeepRecon images were evaluated by American Board of Radiologists certificated physicians." It does not specify the exact number of experts used, nor does it provide details on their years of experience as an example. However, it does state their qualification: American Board of Radiologists certificated physicians.

4. Adjudication Method for the Test Set

The document does not explicitly state the adjudication method (e.g., 2+1, 3+1, none) used for the expert evaluation of the test set. It only mentions that "The evaluation reports from radiologists verified that DeepRecon meets the requirements of clinical diagnosis. All DeepRecon images were rated with equivalent or higher scores in terms of diagnosis quality." This suggests a qualitative review, but the specific consensus method is not detailed.

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

No, a Multi-Reader Multi-Case (MRMC) comparative effectiveness study was not explicitly described in the provided text in terms of quantifying human reader improvement with AI assistance. The expert evaluation focused on whether DeepRecon images met clinical diagnosis requirements and were rated equivalent or higher in quality, rather than measuring a specific effect size of AI assistance on human reader performance.

6. Standalone (Algorithm Only) Performance

Yes, a standalone (algorithm only) performance evaluation was done. The "Acceptance Criteria and Reported Device Performance" table directly shows the performance metrics (Image SNR, Uniformity, Resolution, Contrast, Structure Measurement) of the DeepRecon algorithm itself, compared to images without DeepRecon (NADR). This indicates a standalone assessment of the algorithm's output characteristics.

7. Type of Ground Truth Used for the Test Set

For the quantitative metrics (SNR, uniformity, resolution, contrast, structure measurement), the "ground truth" for comparison appears to be images without DeepRecon (NADR) as a baseline, or potentially direct measurements on those images.

For the qualitative assessment by radiologists, the ground truth was expert opinion/consensus by American Board of Radiologists certificated physicians regarding clinical diagnosis quality.

8. Sample Size for the Training Set

The training set for DeepRecon consisted of data from 264 volunteers. This resulted in a total of 165,837 cases.

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

For the training dataset, the ground truth was established by collecting multiple-averaged images with high-resolution and high SNR. These high-quality images were used as the reference against which the input images (generated by sequentially reducing the SNR and resolution of the ground-truth images) were trained. All data included for training underwent manual quality control.

Ask a specific question about this device