(159 days)
Viz SDH is a notification-only, parallel workflow tool for use by hospital networks and trained clinicians to identify and communicate images of specific patients to a specialist, independent of standard of care workflow.
Viz SDH uses an artificial intelligence algorithm to analyze images for findings suggestive of a prespecified clinical condition and to notify an appropriate medical specialist of these findings in parallel to standard of care image interpretation. Identification of suspected findings is not for diagnostic use beyond notification. Specifically, the device analyzes non-contrast CT images of the head for subdural hemorrhage and sends notifications to a neurovascular or neurosurgical specialist that a suspected subdural hemorrhage has been identified and recommends review of those images can be previewed through a mobile application.
Images that are previewed through the mobile application may be compressed and are for informational purposes only and not intended for diagnostic use beyond notification. Notified clinicians are responsible for viewing non-compressed images on a diagnostic viewer and engaging in appropriate patient evaluation and relevant discussion with a treating physician before making care-related decisions or requests. Viz SDH is limited to analysis of imaging data and should not be used in-lieu of full patient evaluation or relied upon to make or confirm diagnosis.
Viz SDH is a software-only, parallel workflow tool for use by hospital networks and trained clinicians to identify and communicate images of specific patients to an appropriate specialist, such as a neurovascular specialist or neurosurgeon, independent of the standard of care workflow. The system automatically receives and analyses non-contrast CT (NCCT) studies of patients for image features that indicate the presence of a subdural hemorrhage (SDH) using an artificial intelligence algorithm, and upon detection of a suspected SDH, sends a notification so as to alert a specialist clinician of the case.
Viz SDH is a combination of software modules that consists of an image analysis software algorithm and mobile application software module. The Viz SDH image analysis software algorithm is an artificial intelligence machine learning (AI/ML) software algorithm that analyzes non-contrast CT images of the head for a subdural hemorrhage. The Viz SDH Image Analysis Algorithm is hosted on Viz.ai's servers and analyzes applicable stroke-protocoled NCCT images of the head that are acquired on CT scanners and are forwarded to Viz.ai servers. Upon detection of a suspected subdural hemorrhage, the Viz SDH Image Analysis Algorithm sends a notification of the suspected finding.
Viz SDH includes a mobile software module that enables the end user to receive and toggle notifications for suspected subdural hemorrhages identified by the Viz SDH Image Analysis Algorithm. The Viz SDH mobile notification software module is implemented into Viz.ai's nondiagnostic DICOM image viewer, Viz VIEW, which displays CT scans that are sent to Viz.ai's servers. When the Viz SDH mobile notification software module is enabled for a user, the user can receive and toggle the notifications for patients with a suspected subdural hemorrhage, view a unique patient list of patients with a suspected subdural hemorrhage, and view the nondiagnostic CT scan of the patient through the Viz VIEW mobile application. Image viewing through the mobile application interface is for non-diagnostic purposes only.
The Viz SDH device's acceptance criteria and performance are detailed in the provided document.
1. Table of Acceptance Criteria and Reported Device Performance
| Metric | Acceptance Criteria (Lower Bound of 95% CI) | Reported Device Performance |
|---|---|---|
| Sensitivity | 80% | 94% (90% - 97%) |
| Specificity | 80% | 92% (89% - 95%) |
The study also reported an Area Under the Receiver Operating Characteristic Curve (AUC) of 0.96.
The time to notification for SDH was 1.15 ± 0.57 minutes.
2. Sample Size and Data Provenance
- Sample Size for Test Set: 542 non-contrast CT (NCCT) scans (studies).
- Data Provenance: The scans were obtained from three clinical sites in the U.S. The studies included approximately twice as many negative cases as positive cases (66.1% without SDH and 33.9% with SDH).
- Retrospective/Prospective: Not explicitly stated, but the description of "obtained from three clinical sites" and "included in the analysis" typically suggests retrospective data collection for regulatory studies of this nature.
3. Number of Experts and Qualifications for Ground Truth
- Number of Experts: Not explicitly stated.
- Qualifications of Experts: Trained neuro-radiologists.
4. Adjudication Method for the Test Set
The document does not explicitly state the adjudication method used for establishing ground truth from the neuro-radiologists. It only mentions that ground truth was "established by trained neuro-radiologists."
5. Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study
No MRMC comparative effectiveness study was mentioned where human readers' improvement with AI vs. without AI assistance was evaluated. The performance data presented is for the standalone algorithm; however, there is a comparison of the time to notification with a predicate device (Viz ICH) which was previously shown to be clinically meaningful in reducing notification time compared to standard of care.
6. Standalone Algorithm Performance Study
Yes, a standalone performance study was conducted. The sensitivity, specificity, and AUC values reported are for the Viz SDH algorithm without human-in-the-loop performance.
7. Type of Ground Truth Used
The ground truth was established by expert consensus, specifically by trained neuro-radiologists, who compared the Viz SDH's output to the actual NCCT images.
8. Sample Size for the Training Set
The sample size for the training set is not provided in the document. The performance data section focuses entirely on the test set.
9. How 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. It only mentions the approach for the test set.
§ 892.2080 Radiological computer aided triage and notification software.
(a)
Identification. Radiological computer aided triage and notification software is an image processing prescription device intended to aid in prioritization and triage of radiological medical images. The device notifies a designated list of clinicians of the availability of time sensitive radiological medical images for review based on computer aided image analysis of those images performed by the device. The device does not mark, highlight, or direct users' attention to a specific location in the original image. The device does not remove cases from a reading queue. The device operates in parallel with the standard of care, which remains the default option for all cases.(b)
Classification. Class II (special controls). The special controls for this device are:(1) Design verification and validation must include:
(i) A detailed description of the notification and triage algorithms and all underlying image analysis algorithms including, but not limited to, a detailed description of the algorithm inputs and outputs, each major component or block, how the algorithm affects or relates to clinical practice or patient care, and any algorithm limitations.
(ii) A detailed description of pre-specified performance testing protocols and dataset(s) used to assess whether the device will provide effective triage (
e.g., improved time to review of prioritized images for pre-specified clinicians).(iii) Results from performance testing that demonstrate that the device will provide effective triage. The performance assessment must be based on an appropriate measure to estimate the clinical effectiveness. The test dataset must contain sufficient numbers of cases from important cohorts (
e.g., subsets defined by clinically relevant confounders, effect modifiers, associated diseases, and subsets defined by image acquisition characteristics) such that the performance estimates and confidence intervals for these individual subsets can be characterized with the device for the intended use population and imaging equipment.(iv) Stand-alone performance testing protocols and results of the device.
(v) Appropriate software documentation (
e.g., device hazard analysis; software requirements specification document; software design specification document; traceability analysis; description of verification and validation activities including system level test protocol, pass/fail criteria, and results).(2) Labeling must include the following:
(i) A detailed description of the patient population for which the device is indicated for use;
(ii) A detailed description of the intended user and user training that addresses appropriate use protocols for the device;
(iii) Discussion of warnings, precautions, and limitations must include situations in which the device may fail or may not operate at its expected performance level (
e.g., poor image quality for certain subpopulations), as applicable;(iv) A detailed description of compatible imaging hardware, imaging protocols, and requirements for input images;
(v) Device operating instructions; and
(vi) A detailed summary of the performance testing, including: test methods, dataset characteristics, triage effectiveness (
e.g., improved time to review of prioritized images for pre-specified clinicians), diagnostic accuracy of algorithms informing triage decision, and results with associated statistical uncertainty (e.g., confidence intervals), including a summary of subanalyses on case distributions stratified by relevant confounders, such as lesion and organ characteristics, disease stages, and imaging equipment.