(266 days)
HipCheck assists the surgeon to determine quantitative measurements for femoroacetabular impingement (FAI) procedures. HipCheck provides static localization information derived from image processing of intra-operatively acquired static fluoroscopic images, by superposition of virtual measurement tools onto those X-ray images for skeletally mature patients.
HipMap FAI Analysis is a patient-specific report used to support surgeon or radiologist pre-operative clinical decision making. HipMap femoroacetabular impingement (FAI) Analysis provides a morphological analysis of a skeletally mature hip with potential FAI, including measurements and visualizations that describe hip impingement and stability.
HipCheck enables the surgeon to intraoperatively measure alpha angle during hip arthroscopy procedures for femoroacetabular impingement. The software is provided to the user pre-installed on a mobile touchscreen tablet for which it has been tested for compatibility.
Alpha angle is a value used to indicate cam deformity of the femoral head, seen in patients presenting with femoroacetabular impingement. HipCheck provides a visualization tool for surgeons to determine the alpha angle intraoperatively, using virtual measurement tools superimposed on X-ray images collected during the procedure, which informs clinical decision making.
HipCheck is not patient contacting. The user is instructed to appropriately drape the tablet when used in the sterile field.
Stryker HipMap FAI Analysis is a patient-specific report intended for use by surgeons or radiologists to support pre-operative clinical decision making by providing a morphological analysis of a skeletally mature hip with potential femoroacetabular impingement (FAI), including measurements and visualizations that describe hip impingement and stability. HipMap provides three-dimensional analyses, 3D surface reconstructions, and annotated images to support surgeons with pre-operative clinical decision-making.
Here's a summary of the acceptance criteria and study details for the HipCheck device, based on the provided text:
1. Table of Acceptance Criteria and Reported Device Performance
| Acceptance Criteria (Performance Metric) | Reported Device Performance |
|---|---|
| Object Detection AI/ML Model: | |
| Hip presence/absence detection | Automatically detects hip presence/absence (90% Lower Bound of -97.5%) |
| Femur region detection: | |
| - Head center X coordinate accuracy | Within +3.3%/-3.5% |
| - Head center Y coordinate accuracy | Within +3.8%/-4.8% |
| - Neck Angle relative to vertical accuracy | Within +13.63°/-15.35° |
| Mechanical Functionality (leveraged from predicate): | |
| Battery life | Met user needs |
| Tablet weight | Met user needs |
| Tablet securement and attachment force | Evaluates connection between tablet and docking interface |
| User interface temperature and functionality | Functions at operating temperatures |
| RF ablation interference | Met user needs |
| Mounting arm staying force | Met user needs |
| Simulated-use testing | Users successfully used HipCheck as intended |
| Electrical Safety and EMC: | |
| Compliance with IEC 60601-1 | Complies |
| Compliance with IEC 60601-1-2 | Complies |
| Overall Design Validation (HipCheck): | Users successfully used HipCheck as intended to determine alpha angle and utilize tools. |
2. Sample Size Used for the Test Set and Data Provenance
- Test Set Sample Size: 745 fluoroscopic images.
- 184 images: Did not contain images of hips, used to test false positive detection.
- 561 images: From 81 hips.
- Data Provenance: Geographically, images came from 6 clinical sites in the United States, Netherlands, and Germany. The images were collected during product development cadaver labs or from anonymized log files from patients undergoing surgery. This data appears to be a mix of prospective (cadaver labs) and retrospective (anonymized log files) sources.
3. Number of Experts Used to Establish the Ground Truth for the Test Set and Qualifications of Those Experts
- Number of "Experts" (Taggers): Two people.
- Qualifications of Experts: They were "trained to use the software" for labeling the femur with the precise location of the femoral head and neck. Specific professional qualifications (e.g., radiologist, orthopedist) or years of experience are not specified in the provided text.
4. Adjudication Method for the Test Set
- For the object detection AI/ML model, testing was done against the average value of the two taggers. This implies a form of consensus or averaging for ground truth.
5. Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study
- No MRMC comparative effectiveness study involving human readers with and without AI assistance is explicitly described for the HipCheck Alpha Angle algorithm.
- However, for the HipMap FAI Analysis (a component of the HipCheck device), a segmentation accuracy and reliability study was conducted. This study reviewed:
- "Performance of Stryker personnel segmenting pelvic CT scans using HipMap workflow software against trained third-party personnel performing image segmentation of the same scans using 510(k) cleared software."
- "Reliability of segmentation between Stryker personnel (inter-rater reliability)."
- "Reliability of the HipMap FAI Analysis by comparing clinical measurement outputs generated from the third-party segmentation (external rater vs internal rater), Stryker employee segmentations (inter-rater reliability), and iterations of segmentations performed by the same Stryker employee (intra-rater reliability)."
- This is a comparative study, but it's focused on segmentation accuracy and reliability between different personnel and software, rather than the "human readers + AI vs. human readers alone" paradigm. Effect sizes are not mentioned in the provided text for this comparison.
6. Standalone Performance (Algorithm Only) Study
- Yes, a standalone performance testing was conducted for the object detection AI/ML model, which is part of the HipCheck device's image processing pipeline. The results are detailed in the table above (90% Lower Bound of -97.5% for detection, and percentage/degree accuracies for coordinate and angle measurements).
7. Type of Ground Truth Used
- For the object detection AI/ML model's standalone performance testing: Expert Consensus/Annotation (labeled by two trained individuals, with the average value used as ground truth).
8. Sample Size for the Training Set
- The sample size for the training set is not explicitly stated in the provided text. It only mentions that the test dataset was "independent of the data used during model training."
9. How the Ground Truth for the Training Set Was Established
- The method for establishing ground truth for the training set is not explicitly stated in the provided text. It can be inferred that it likely followed a similar annotation process to the test set, but specific details are absent.
§ 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).