(232 days)
OrthoCAD is an option that provides the morphometry of the lumbo-sacral section of the spine, by means of semi-automatic segmentation of MR images, the generation of the relative 3D model and calculation of the significant geometrical properties of the vertebral bodies and spinal canal. When this data is interpreted by a trained physician, it can yield information that may assist diagnosis.
The OrthoCAD software option is a software package intended to be used with Esaote Gscan system cleared via K111803. OrthoCAD provides the morphometry of the lumbosacral section of the spine, by means of semi-automatic segmentation of MR images, the generation of the relative 3D model and calculation of the significant geometrical properties of the vertebral bodies and spinal canal. When this data is interpreted by a trained physician, it can yield information that may assist diagnosis.
G-scan is a Magnetic Resonance (MR) system, which produces images of the internal structures of the patient's limbs and joints.
The OrthoCAD system allows you to visualize, analyse and compare Magnetic Resonance images. The system is connected to a database that enables the physician to keep track of all the patients examined over time.
The images are acquired by running FSE T2 Rel sequences on the G-scan of the lumbosacral tract of the vertebral spine, in the sagittal plane, and are transferred to the OrthoCAD database following acquisition.
When the MR images are stored on the OrthoCAD database, the user can proceed with a manual or semi-automatic (wizard) segmentation of the vertebral bodies (from L1 up to S1) and of the spinal canal. During the segmentation of the anatomical elements, 3D models are constructed based on the segmented structures.
When the segmentation procedure has been terminated, the user can proceed with the evaluation of the following significant clinical parameters:
- . Vertebral wedging
- Listhesis index .
- Intervertebral translation index .
- Intervertebral angles .
- Vertebral collapse index .
- . Spinal curvature
- . Spinal canal thickness
- . Spinal canal section
- Foramen area . .
Following this process, the endoscopic virtual navigation within the segmented spinal canal is enabled.
Finally, if the user has worked on MR images acquired both in the clinostatic and orthostatic mode, the measures calculated and the virtual navigation of the two the positions can be compared, and a report containing all the information is produced.
OrthoCAD is made up of six environments:
- . Patient Management: contains the functions required for the display and management of patients stored in the database associated with the system.
- Home: keeps track of the procedures executed overtime for the selected patient . (analyses present, status of examinations associated with the various analyses, etc.).
- . Segmentation: carries out the functions used for the segmentation and those related to the construction of 3D models of anatomic elements
- Measurements: includes all tools required to measure the clinical parameters . used for the analysis of the currently selected exam.
- Navigation: enables endoscopic virtual navigation within the segmented . anatomical structures by means of the definition of anatomic points, in order to construct one (or more) navigation routes.
- . Comparison: enables the comparison of two different examinations within the same analysis or within different analyses provided they are the same type. This environment enables:
- . The simultaneous display, or superimposed display when required, of anatomical elements which belong to the two volumes being compared.
- 미 The simultaneous display of the different measurements, with an indication of the main differences between these values.
Here's an analysis of the provided 510(k) summary for the OrthoCAD software option, detailing the acceptance criteria and study information:
1. Acceptance Criteria and Reported Device Performance
The provided document does not explicitly state quantitative acceptance criteria (e.g., a specific percentage accuracy or precision threshold). Instead, it describes a series of comparative tests that were performed to demonstrate the device's performance against manual methods. The reported performance indicates that these tests were conducted, and the summary's conclusion is that the device "met performance requirements and is as safe and effective as the predicate devices."
Based on the "Tests performed" section, the implicit acceptance criteria are related to the consistency, variability, repeatability, reproducibility, and correctness of measurements when compared to manual segmentation and measurement.
| Acceptance Criteria (Inferred from "Tests performed") | Reported Device Performance |
|---|---|
| Comparison between manual and semi-automated segmentation on lumbar and first sacral vertebrae. | Tests performed; device "met performance requirements." |
| Comparison between manual and semi-automated segmentation on spinal canal. | Tests performed; device "met performance requirements." |
| Comparison of manual and semi-automated measurements and evaluation of variability, repeatability, and reproducibility. | Tests performed; device "met performance requirements." |
| Validation of new software OrthoCAD in its correctness in measuring MRI images of the Lumbar spine. | Tests performed; device "met performance requirements." |
2. Sample Size Used for the Test Set and Data Provenance
The document does not specify the sample size used for the test set (number of images, cases, or segmented structures).
The data provenance is implicitly from MRI images of the lumbosacral tract of the vertebral spine acquired using an Esaote G-scan system. The country of origin of the data is not explicitly stated but can be inferred to be associated with Esaote S.p.A., an Italian company. The document does not specify if the data was retrospective or prospective.
3. Number of Experts Used to Establish Ground Truth and Qualifications
The document does not explicitly state the number of experts used to establish the ground truth for the test set, nor does it detail their specific qualifications (e.g., radiologist with X years of experience). It refers to the "manual segmentation" as the comparative method, implying that experts performed this manual segmentation, but the specifics are absent.
4. Adjudication Method for the Test Set
The document does not specify an adjudication method (e.g., 2+1, 3+1, none) for the test set. The comparison is described as being between "manual and semi-automated segmentation" and "manual and semi-automated measurements," suggesting that the manual method served as the reference, but how disagreements or variations in the manual process were handled is not detailed.
5. Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study
The document does not mention or describe a multi-reader multi-case (MRMC) comparative effectiveness study. The focus is on comparing the semi-automated software's performance against manual methods, rather than assessing the improvement of human readers with AI assistance.
6. Standalone Performance Study
Yes, a standalone performance study was implicitly done. The "Tests performed" section describes comparisons between the semi-automated segmentation and measurements (algorithm only) and manual methods. This indicates that the algorithm's performance was evaluated independently from a human-in-the-loop scenario, by comparing its outputs directly to the manually established ground truth.
7. Type of Ground Truth Used
The type of ground truth used was expert consensus / manual segmentation and measurement. The tests involved a "comparison between the manual and semi-automated segmentation" and "comparison of the manual and semi-automated measurements," indicating that the gold standard was derived from manual interpretation and manipulation by human experts.
8. Sample Size for the Training Set
The document does not specify the sample size used for the training set.
9. How the Ground Truth for the Training Set Was Established
The document does not provide information on how the ground truth for the training set was established. In fact, it doesn't mention a distinct training set at all, which is common for older 510(k) submissions, especially for software that is more rules-based or semi-automatic rather than deep learning-based. It's possible the "manual segmentation" and "measurements" used for testing were also part of the development and refinement process, but a separate, explicitly defined training set with its ground truth establishment is not detailed.
§ 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).