(225 days)
MySegmenter (v2.0.0) is a medical images segmentation tool that converts CT and MRI scans into 3D anatomical models. The models are intended for surgical planning or educational purposes but are not to be used in the operating room.
The software must be used under the supervision of qualified medical professionals. The 3D models and replicas are to be utilized solely for pre-operational planning in orthopedic and craniomaxillofacial cases, and for educational purposes; they must not be used in the operating room. The MR images have not been tested for craniomaxillofacial applications.
MySegmenter (v2.0.0) is a software platform for visualizing medical images in 3D mesh format. processes images by analyzing the luminance of individual pixels to segment and highlight the regions of interest (specific anatomical areas). MySegmenter supports editing, modification and manipulation of segmented areas using various segmentation tools. The software features include:
- Importing medical images in DICOM and other formats.
- Viewing and navigating through DICOM images.
- Segmenting selected regions using generic tools (segments, island effects, Boolean operations, etc.)
- Editing segments with multiple slice edits, fast marching, watershed effects, etc.
- Generating editable 3D STL files for different applications.
- Converting the segment to mesh models.
- Exporting the segmented models in STL format suitable for further manipulation and manufacturing.
Here's a breakdown of the acceptance criteria and study information for MySegmenter (v2.0.0), based on the provided FDA 510(k) clearance letter and summary:
Acceptance Criteria and Reported Device Performance
| Metric | Acceptance Criteria (Implicit) | Reported Device Performance |
|---|---|---|
| Digital Model Accuracy (CT & MRI) | DICE and Jaccard similarity coefficients, and Hausdorff distance (for surface accuracy) comparable to predicate device. Hausdorff distance below 1 mm. | DICE and Jaccard similarity coefficients: 90% to 98% (depending on manual editing). Hausdorff distance: below 0.6 mm. |
| 3D Print Accuracy | Geometric accuracy (surface-to-surface comparison via Hausdorff distance) below 1 mm compared to digital models from the predicate device. | Hausdorff distance: consistently below 1 mm. |
Study Details
2. Sample sizes used for the test set and data provenance
Test Set - CT and MRI Accuracy Studies:
- Sample Size: 30 datasets (15 CT, 15 MRI)
- Data Provenance: Indian origin, evenly distributed between adult male and female subjects.
- Retrospective/Prospective: Not explicitly stated, but typically such studies utilize retrospective data sets.
Test Set - 3D Printing Accuracy:
- Sample Size: 20 datasets (15 CT, 5 MRI)
- Data Provenance: Indian origin, evenly distributed between adult male and female subjects.
- Retrospective/Prospective: Not explicitly stated, but typically such studies utilize retrospective data sets.
3. Number of experts used to establish the ground truth for the test set and the qualifications of those experts
- Number of Experts: Not specified. The document states "experts in the field of medical imaging."
- Qualifications of Experts: "Experts in the field of medical imaging." Specific experience levels or specializations (e.g., radiologist with X years of experience) are not provided.
4. Adjudication method for the test set
- Adjudication Method: Inter-expert agreement was evaluated using Intraclass Correlation Coefficient (ICC) scores to ensure absolute agreement for the ground truth models created by experts. This implies that multiple experts reviewed and agreed upon the reference models.
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
- MRMC Study: No, an MRMC comparative effectiveness study involving human readers and AI assistance was not explicitly mentioned or described. The studies focused on the accuracy of the device's output (3D models and prints) compared to a predicate device and expert-created ground truth.
6. If a standalone (i.e. algorithm only, without human-in-the-loop performance) was done
- Standalone Performance: Yes, the accuracy studies (CT and MRI Accuracy Studies, and 3D Printing Accuracy) describe the performance of the MySegmenter (v2.0.0) algorithm in generating 3D models and prints, which are then compared to ground truth or predicate device outputs. While manual editing is mentioned as influencing performance for DICE/Jaccard, the core evaluation is on the software's ability to produce these models. The device description also highlights "Manual and Semiautomatic tools for segmentation," suggesting that the algorithm performs segmentation which users can then refine.
7. The type of ground truth used (expert consensus, pathology, outcomes data, etc.)
- Type of Ground Truth: For the CT and MRI Accuracy Studies, the ground truth was expert-created reference models using the predicate device. This is a form of expert consensus or expert-derived reference. For the 3D Printing Accuracy, the digital models produced by the predicate device served as the comparison point for the 3D printed physical models.
8. The sample size for the training set
- Training Set Sample Size: Not provided in the document. The document only details the test set sizes for validation.
9. How the ground truth for the training set was established
- Training Set Ground Truth Establishment: Not provided in the document, as the training set details are not mentioned.
§ 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).