(77 days)
Accipiolx is a software workflow tool designed to aid in prioritizing the clinical assessment of adult non-contrast head CT cases with features suggestive of acute intracranial hemorrhage in the acute care environment. Accipiolx analyzes cases using an artificial intelligence algorithm to identify suspected findings. It makes case-level output available to a PACS/workstation for worklist prioritization or triage.
Accipiolx is not intended to direct attention to specific portions of an image or to anomalies other than acute intracranial hemorrhage. Its results are not intended to be used on a stand-alone basis for clinical decision-making nor is it intended to rule out hemorrhage or otherwise preclude clinical assessment of CT cases.
Accipiolx is a software device designed to be installed within healthcare facility radiology networks to identify and prioritize non-contrast head CT (NCCT) scans based on algorithmically-identified findings of acute intracranial hemorrhage (alCH). The device, developed using computer vision and deep learning technologies, facilitates prioritization of CT scans containing findings of alCH. There are two main components of the software device: (1) the Accipiolx Agent and (2) the MaxQ-Al Engine. The Agent serves as an active conduit which receives head CT studies from a PACS and transfers them to the Engine. After successful processing of a case via the MaxQ-Al Engine, the Accipiolx Agent receives the Engine results and returns them to the PACS or workstation for use in worklist prioritization.
Accipiolx works in parallel to and in conjunction with the standard care of workflow. After a CT scan has been performed, a copy of the study is automatically retrieved and processed by Accipiolx. The device performs identification and classification of objects consistent with alCH, and provides a case-level indicator which facilitates prioritization of cases with potential acute hemorrhagic findings for urgent review.
Here's an analysis of the acceptance criteria and study as described in the provided text:
1. Table of Acceptance Criteria and Reported Device Performance
| Metric | Acceptance Criteria (Predefined Performance Goals) | Reported Device Performance |
|---|---|---|
| Sensitivity | Less than 92% (inferred from "exceeded" statement) | 92% (95% Cl: 87.29-95.68%) |
| Specificity | Less than 86% (inferred from "exceeded" statement) | 86% (95% Cl: 80.18-90.81%) |
Notes: The document states that the reported results "exceeded the predefined performance goals." This implies the acceptance criteria were less than the achieved performance, meaning the device had to perform at least
as well as a certain threshold. However, the exact numerical thresholds for the acceptance criteria are not explicitly stated, so I've inferred them based on the "exceeded" statement.
2. Sample size used for the test set and the data provenance
- Sample Size: 360 cases
- Data Provenance:
- Country of Origin: Not explicitly stated, but collected from "over 30 US sites." This suggests the data originated from the United States.
- Retrospective/Prospective: Retrospective study.
3. Number of experts used to establish the ground truth for the test set and the qualifications of those experts
- Number of Experts: At least two expert neuroradiologist readers.
- Qualifications of Experts: Expert neuroradiologist readers. No specific experience in years is provided.
4. Adjudication method for the test set
- Adjudication Method: Concurrence of at least two expert neuroradiologist readers. This implies a 2-reader consensus model.
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
- A MRMC comparative effectiveness study involving human readers with vs. without AI assistance was not done. The study described focused on the standalone performance of the AI algorithm.
6. If a standalone (i.e. algorithm only without human-in-the-loop performance) was done
- Yes, a standalone study was done. The performance testing specifically evaluated the "device sensitivity and specificity...compared to ground truth." This describes the algorithm's performance in isolation.
7. The type of ground truth used
- Type of Ground Truth: Expert consensus (established by concurrence of at least two expert neuroradiologist readers).
8. The sample size for the training set
- Sample Size for Training Set: Not explicitly stated. The text mentions "Accipiolx was developed using a training CT cases collected from multiple institutions and CT manufacturers," but it does not provide a specific number for the training set size.
9. How the ground truth for the training set was established
- How Ground Truth for Training Set was Established: Not explicitly stated how the ground truth for the training specific was established. The document mentions "optimization of object and feature identification, algorithmic training and selection/optimization of thresholds," which implies a process was followed, but the method of establishing training ground truth is not detailed.
§ 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.