(90 days)
The Viatronix V3D Colon is a system for the display and visualization of 3D and 2D medical image data of the colon derived from DICOM 3.0 compliant CT and MR scans, for the purpose of patient screening for detection of colon cancers, polyps, masses, and other lesions. It provides functionality for display, measurement and electronic cleansing to assure complete visualization of the colon from rectum to cecum and vice versa for both prone and supine views. It generates a centerline for guided interactive navigation and fly through of the entire colon, and also includes a reporting facility to enhance workflow for patient screening. It is intended for use by Radiologists, Clinicians and referring Physicians to process, render, review, archive print and distribute colon image studies utilizing PC hardware.
The Viatronix V3D Colon contains all of the required hardware and software components to provide interactive 3D and 2D views of diagnostic CT and MR scan images of the colon. The views include both inner and outer surface 3D volume rendered images as well as orthogonal and multiplaner reformatted 2D images. This ability to view the dataset in different perspectives from which it was acquired is performed by first transferring DICOM images from the MR or CT scanners to the Viatronix V3D Colon, which automatically identifies regions of interest and displays thesc regions to the user in the above mentioned views. The user can then navigate freely within the dataset/region of interest or follow automatically computed paths to fly through the colon or around the outside of the colon structure. Measurements of the size of colon polyps, masses or lesion can be made for patient screening and for planning treatment.
Here's a breakdown of the acceptance criteria and study information for the Viatronix V3D Colon, revision 1.3, based on the provided document:
Acceptance Criteria and Device Performance
| Acceptance Criteria | Reported Device Performance |
|---|---|
| Preamble Goals | |
| Accurate 3D model of the entire colon is generated | Verified |
| Fly-through in the 3D model for the entire colon is possible | Verified |
| Viewing 2D and MPR images of the entire colon is possible | Verified |
| Measurements are accurate (using phantom) | Verified |
| Ability to fly through and visualize the entire colon | Demonstrated |
| Ability to visualize and measure polyps, masses and lesions for patient screening to detect colon cancers | Demonstrated |
| Clinical Study Performance (Colorectal Neoplasia Screening in Asymptomatic Adults) | |
| Virtual Colonoscopy (VC) Sensitivity for 10mm adenomatous polyps | 93.8% |
| Virtual Colonoscopy (VC) Sensitivity for 8mm adenomatous polyps | 93.9% |
| Virtual Colonoscopy (VC) Sensitivity for 6mm adenomatous polyps | 88.7% |
| Virtual Colonoscopy (VC) Specificity for 10mm adenomatous polyps | 96.0% |
| Virtual Colonoscopy (VC) Specificity for 8mm adenomatous polyps | 92.2% |
| Virtual Colonoscopy (VC) Specificity for 6mm adenomatous polyps | 79.6% |
| Detection of malignant polyps (2 total) | Both detected on VC |
Note on Acceptance Criteria: The document primarily lists "goals of clinical testing" rather than explicit numerical acceptance criteria for all aspects. For sensitivity and specificity, the reported values can be considered the demonstrated performance against assumed implicit acceptance thresholds for a diagnostic tool in this context.
Study Details
1. Sample Size Used for the Test Set and Data Provenance:
- Sample Size: A total of 1233 asymptomatic adults.
- Data Provenance: US Government's Department of Defense Medical Centers (United States). The study was prospective, as it involved subjects undergoing both virtual colonoscopy and optical colonoscopy on the same day.
2. Number of Experts Used to Establish the Ground Truth for the Test Set and Qualifications:
- The document implies that the "Results" section of the major clinical study (8.1) references data where Optical Colonoscopy (OC) served as a comparator for sensitivity and specificity. However, it does not explicitly state the number of experts who established the ground truth from OC, or their qualifications.
- For the "Initial Clinical tests" (8.2), a single radiologist reviewed the clinical data.
- Qualifications of Radiologist (8.2): Experienced with predicate devices.
3. Adjudication Method for the Test Set:
- For the main clinical study (8.1), the document mentions a comparison with optical colonoscopy. One paragraph states: "Two polyps were malignant; both were detected on VC and one was missed on OC before the results on VC were revealed." This implies a potential sequential or blinded review process where OC findings were known post-VC review, but a formal adjudication method (like 2+1) is not explicitly detailed.
- For the "Initial Clinical tests" (8.2), the review was performed by a single radiologist, suggesting no formal adjudication.
4. If a Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study was done, and its effect size:
- No, a formal MRMC comparative effectiveness study aiming to measure the effect size of human readers with vs. without AI assistance is not described.
- The main clinical study (8.1) compares the device's performance (VC) directly against Optical Colonoscopy (OC) in terms of sensitivity and specificity, effectively comparing the AI-assisted diagnostic pipeline (VC reviewed by an unnamed number of readers) to the gold standard (OC). It does not isolate the improvement from AI assistance to a human reader.
5. If a Standalone (i.e., algorithm only without human-in-the-loop performance) was done:
- No, the studies described do not appear to be standalone algorithm-only performance evaluations.
- The device is a "system for the display and visualization" intended for use by "Radiologists, Clinicians and referring Physicians." The clinical study evaluated "Virtual Colonoscopy (VC) ... by patient," which implicitly involves human interpretation of the V3D Colon output. The "Initial Clinical tests" section explicitly states that a radiologist reviewed the rendering.
6. The Type of Ground Truth Used:
- For the main clinical study (8.1): Optical Colonoscopy (OC) findings were used as the primary ground truth. This is a form of clinical/procedural ground truth.
- For the non-clinical tests (7): Phantoms with structures of known size and distance were used. This is a form of physical phantom ground truth.
7. The Sample Size for the Training Set:
- The document does not specify a sample size for the training set. It discusses the development process ("developed in a manner consistent with accepted standards for software development, including test protocols") and testing, but not the data used to train the AI/algorithm component of the system.
8. How the Ground Truth for the Training Set was Established:
- The document does not provide information on how ground truth was established for any training set, as a training set is not explicitly discussed.
§ 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).