(95 days)
The Zebra Health.JOINT device is a software tool for 3D reconstruction of bones from a set of 2D radiographs. The device is intended for assisting clinicians in the preoperative planning of knee orthopedic surgical procedures. Zebra's HealthJOINT analyzes cases using an artificial intelligence algorithm for the 3D model reconstruction. In addition to the model, the software provides a list of anatomical landmarks with their position on the 3D model. The result is made available via a 3rd parties' software interface for further display and analysis of the 3D bone model. Clinical judgement and experience are required to properly use the models produced by this software.
Zebra's HealthJOINT device is a software product that uses an artificial intelligence algorithm to analyze X-ray scans. The HealthJOINT is indicated for the analysis of X-rays scans. The device receives a set of 2D radiographs and automatically provides a 3D model of the bones together with a list of anatomical landmarks with their position on the 3D model may be used by physicians for pre-operative planning of knee orthopedic surgeries. The HealthJoint supports 3D reconstructions of healthy bones, and osteoarthritis patients graded 1 to 4 based on the Kellgren-Lawrence grading system.
The HealthJOINT device functions as a component that can be used by 3rd parties via an API to generate the 3D models and provides a list of anatomical landmarks with their position on the 3D model. The software communicates with the API only, and is not user-facing. The software does not recommend clinical decisions or treatment.
The software is intended for use by clinicians in conjunction with additional patient information and professional judgment.
The following modules compose the HealthJoint software:
Data input and validation: performs validation of the input, X-ray DICOM images, assesses the input data (i.e. age, modality, view) to ensure compatibility for processing by the algorithm.
HealthJoint algorithm: Once the study has been validated the algorithm analyzes the AP (anteriorposterior) along with the LAT (lateral) knee X-ray study in order to provide 3D bone models and locations of anatomic landmarks.
IMA Integration feature: provides the capability to post studies for processing, get the study analysis status and the results of successful study analysis via a Web API.
Error codes feature: In the case of a study failure during data validation or the analysis by the algorithm, an error is provided to the calling 3rd party via the Web API.
1. Acceptance Criteria and Reported Device Performance
| Metric | Acceptance Criteria | Reported Device Performance |
|---|---|---|
| 3D Model Accuracy (RMSE) | Not explicitly stated as a numerical threshold, but implies "met the pre-defined success criteria" | |
| Femur (RMSE) | - | 1.14 (95% CI: [1.097, 1.187]) |
| Tibia (RMSE) | - | 1.05 (95% CI: [1.005, 1.087]) |
| Fibula (RMSE) | - | 0.94 (95% CI: [0.891, 0.986]) |
| Anatomical Landmark Positioning (Precision) | "Met the performance goal" (not explicitly stated as a numerical threshold) | Standard deviation of the distance (specific values not provided in the summary, but stated as meeting the goal) |
2. Sample Size and Data Provenance for Test Set
- Sample Size: 67 pairs of knee X-rays and CT scans.
- Data Provenance: The document states "US Board-Certified radiologist" when describing the ground truth expert, but doesn't explicitly state the country of origin for the data itself. The study was retrospective.
3. Number and Qualifications of Experts for Ground Truth (Test Set)
- Number of Experts: One
- Qualifications of Expert: Experienced US Board-Certified radiologist.
4. Adjudication Method (Test Set)
The document does not explicitly describe an adjudication method for the test set. It mentions ground truth established by a single experienced US Board-Certified radiologist.
5. Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study
No, a multi-reader multi-case (MRMC) comparative effectiveness study was not conducted. The performance evaluation was a standalone retrospective study of the device's accuracy.
6. Standalone Performance Study
Yes, a standalone performance study was conducted. The document states: "The HealthJOINT device performance was evaluated in a stand-alone retrospective study for accuracy..."
7. Type of Ground Truth Used (Test Set)
The ground truth for the 3D model accuracy and anatomical landmark positioning was established by comparing the software's output to an "established Ground Truth by an experienced US Board-Certified radiologist" using CT scans as a reference.
8. Sample Size for Training Set
The document does not provide the sample size used for the training set. It focuses on the performance evaluation using an independent test set.
9. How Ground Truth for Training Set Was Established
The document does not provide information on how the ground truth for the training set was established.
§ 892.2050 Medical image management and processing system.
(a)
Identification. A medical image management and processing system is a device that provides one or more capabilities relating to the review and digital processing of medical images for the purposes of interpretation by a trained practitioner of disease detection, diagnosis, or patient management. The software components may provide advanced or complex image processing functions for image manipulation, enhancement, or quantification that are intended for use in the interpretation and analysis of medical images. Advanced image manipulation functions may include image segmentation, multimodality image registration, or 3D visualization. Complex quantitative functions may include semi-automated measurements or time-series measurements.(b)
Classification. Class II (special controls; voluntary standards—Digital Imaging and Communications in Medicine (DICOM) Std., Joint Photographic Experts Group (JPEG) Std., Society of Motion Picture and Television Engineers (SMPTE) Test Pattern).