(180 days)
The SafeCT-29 is intended for provided Tomography Dose Check feature to Computed Tomography X-ray systems.
The SafeCT-29 is specifically indicated for providing the Computed Tomography Dose Check feature which notifies and alerts the CT equipment operators, prior to a scan, if the estimated dose index is above the predefined thresholds, for CT scanners not equipped with this functionality. The device is indicated for use by professional personnel.
Medic Vision's SafeCT-29 provides a vendor-neutral Radiation Dose Check functionality, in accordance with the Dose Check feature that is specified by the NEMA XR-25 Standard. The device is a software and hardware system which includes Computer, dedicated display and controls. The SafeCT-29 is interfaced to CT scanners that are not equipped with the Dose Check function. The device is connected to the CT Console video output via a standard video connector. The SafeCT-29 Computer's internal video splitter captures the CT Console display video, in real time, without affecting neither the CT console itself nor its display. This analog video signal is digitized by a video grabber. The SafeCT-29 software analyzes the digital input video using OCR software continuously. The radiation dose information as calculated by the scanner and displayed to the CT operator is extracted and analyzed in real time. The SafeCT-29 captures the video of the CT Console display in real time, without affecting neither the CT console itself nor its display and workflow. As specified by NEMA XR-25 Standard, the safeCT-29 notifies and alerts the CT operators, prior to a scan, if the estimated dose index is above the predefined values set by the operating group, practice, or organization. SafeCT-29 prevents continuing an over-the-limit scan unless dose levels are reconfirmed by the operator, in accordance with NEMA XR-29*.The device records the details of such events, including the operator details and decisions. This record is available to the user.
Here's a breakdown of the acceptance criteria and the study proving the device meets them, based on the provided text:
Acceptance Criteria and Device Performance
| Acceptance Criteria | Reported Device Performance |
|---|---|
| Specificity in detecting correct input data (protocol and dose data) | Better than 99.4% |
| Delay in detecting correct input data | Less than 300ms |
| Accuracy of accumulated CTDIvol calculation | Follows AAPM guidelines, assumes "worst-case scenario" (calculated CTDIvol may be higher than actual value) to prevent over-the-limit scans. |
| Compliance with NEMA XR-25 Standard (Dose Check features) | The software output meets the input as defined by the Software Requirements Specifications (SRS), which encompasses Dose Check features as specified by NEMA XR-25 standard, display & operation controls, and applicable risk mitigations. |
| Performance in simulating real-world CT scanner input | The device processed recorded data successfully and its Dose Check functionality was tested, confirming it meets performance specifications. |
| Usability and reliability in user environment | Verified at 4 imaging centers using 4 different CT scanner models by end-users and Medic Vision engineers. |
| Conformity to defined user needs and intended uses | Confirmed through validation testing of production units at 5 medical centers with 5 different CT scanner models, covering installation, operation, functionality processes, and features. |
Study Details
-
Sample sizes used for the test set and the data provenance:
- Test Set (for OCR software verification): 1800 screen images captured from various CT scanners. These images contained more than 6000 numeric and alphanumeric text images.
- Test Set (for system performance verification): Sequence of data recorded from various CT scanners (specific number not provided, but implies a diverse set).
- Test Set (for usability and reliability): Conducted at 4 imaging centers using 4 different CT scanner models.
- Test Set (for validation): Conducted at 5 medical centers with 5 different CT scanner models.
- Data Provenance: The text states "various CT scanners" and "different CT scanner models," indicating a diverse, but unspecified, range. It does not explicitly state the country of origin or whether the data was retrospective or prospective. Given the nature of a product submission, it's likely retrospective data collection from existing scanners.
-
Number of experts used to establish the ground truth for the test set and the qualifications of those experts:
- The text does not specify the number or qualifications of experts used to establish the ground truth. For the OCR performance, the "results were compared to the original CT screens' data," implying a direct comparison to the source rather than expert interpretation. For the overall system performance, the "results of process confirmed that device meets its performance specifications," implying validation against predefined expectations rather than expert consensus on individual cases.
-
Adjudication method (e.g., 2+1, 3+1, none) for the test set:
- The text does not describe an adjudication method for the test set. For the OCR performance, it explicitly states a direct comparison ("results were compared to the original CT screens' data"). For other tests, it implies automated or predefined checks against specifications.
-
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:
- No, an MRMC comparative effectiveness study was not done. The device, SafeCT-29, is an add-on to CT scanners to provide a "Dose Check" feature. It functions by analyzing the CT console's video output via OCR to alert operators about high dose indexes prior to a scan. It does not involve human readers interpreting images, nor does it provide AI assistance for image interpretation. Therefore, there's no "human readers improve with AI" metric applicable here.
-
If a standalone (i.e., algorithm only without human-in-the-loop performance) was done:
- Yes, a standalone performance evaluation of the OCR software was done. The "accuracy of the OCR software" was tested by inputting 1800 screen images and comparing its extracted data (dose levels and protocols) to the "original CT screens' data." This can be considered a standalone evaluation of the core OCR algorithm. The overall system's functionality was also verified by simulating inputs and testing the "Dose Check functionality," which is also a standalone assessment of the algorithm's performance against specifications.
-
The type of ground truth used (expert consensus, pathology, outcomes data, etc.):
- For OCR software: The ground truth was the original data displayed on the CT screens from which the images were captured. This is a direct, objective comparison.
- For overall system performance: The ground truth was based on predefined Software Requirements Specifications (SRS), which were aligned with the NEMA XR-25 standard and AAPM guidelines for dose calculation. This indicates a rules-based and standard-compliant ground truth.
-
The sample size for the training set:
- The document does not provide information on the sample size used for training the OCR software or any other part of the SafeCT-29 system. It focuses solely on verification and validation (testing) activities. It mentions an "Off-The Shelf (OTS) OCR software" is embedded, implying that Medic Vision may not have trained the core OCR algorithm from scratch, but rather verified its performance within their system.
-
How the ground truth for the training set was established:
- As no information on a training set is provided, the method for establishing its ground truth is also not described. If an OTS OCR was used, its training would have been handled by its original developer. For Medic Vision's own development, if any training was involved, the method is not disclosed.
§ 892.1750 Computed tomography x-ray system.
(a)
Identification. A computed tomography x-ray system is a diagnostic x-ray system intended to produce cross-sectional images of the body by computer reconstruction of x-ray transmission data from the same axial plane taken at different angles. This generic type of device may include signal analysis and display equipment, patient and equipment supports, component parts, and accessories.(b)
Classification. Class II.