Cina is a radiological computer aided triage and notification software in the analysis of (1) not-enhanced head CT images and (2) CT angiography of the head.

The device is intended to assist hospital networks and trained radiologists in workflow triage by flagging and communicating suspected positive findings of (1) head CT images for Intracranial Hemorthage (ICH) and (2) head CT angiography for large vessel occlusion (LVO) of the anterior circulation (distal ICA, MCA-M1 or proximal MCA-M2). Cina uses an artificial intelligence algorithm to analyze images and highlight cases with detected (1) ICH or (2) LVO on a standalone Web application in parallel to the ongoing standard of care image interpretation. The user is presented with notifications for cases with suspected ICH or LVO findings.

Notifications include compressed preview images that are meant for informational purposes only, and are not intended for diagnostic use beyond notification. The device does not alter the original medical image, and it is not intended to be used as a diagnostic device.

The results of Cina are intended to be used in conjunction with other patient information and based on professional judgement to assist with triage/prioritization of medical images. Notified clinicians are ultimately reviewing full images per the standard of care.

Device Description

Cina is a radiological computer-assisted triage and notification software device.

The software system is based on algorithm-programmed components and is comprised of a standard off-the-shelf operating system and additional image processing applications.

DICOM images are received, recorded and filtered before processing. The series are processed chronologically by running algorithms on each series to detect suspicious results of an intracranial hemorrhage (ICH) or a large vessel occlusion (LVO), then notifications on the flagged series are sent to the Worklist Application.

The Worklist Application (on premise) displays the pop-up notifications of new studies with suspected findings when they come in, and provides both active and passive notifications. Active notifications are in the form of a small pop-up containing patient name, accession number and the type of suspected findings (ICH or LVO). All the non-enhanced head CT images and head CT angiography studies received by Cina device are displayed in the worklist and those on which the algorithms have detected a suspected finding (ICH or LVO) are marked with an icon (i.e., passive notification). In addition, a compressed, small black and white image that is marked "not for diagnostic use" is displayed as a preview function. This compressed preview is meant for informational purposes only, does not contain any marking of the findings, and is not intended for primary diagnosis beyond notification. Presenting the radiologist with notification facilitates earlier triage by allowing one to prioritize images in the PACS. Thus, the suspect case receives attention earlier than would have been the case in the standard of care practice alone.

AI/ML Overview

Here's a breakdown of the acceptance criteria and study details for the Cina device, based on the provided text:

1. Table of Acceptance Criteria and Reported Device Performance

Performance Metric	Acceptance Criteria (Performance Goal)	Cina-ICH Reported Performance	Cina-LVO Reported Performance
Sensitivity	80%	91.4% (95% Cl: 87.2% – 94.5%)	97.9% (95% Cl: 94.6% - 99.4%)
Specificity	80%	97.5% (95.8% – 98.6%)	97.6% (95% Cl: 95.1% - 99%)
AUC (ROC)	Not explicitly stated, but comparable to predicate	0.94	0.98
Overall Agreement (Accuracy)	Not explicitly stated	95.6%	97.7%
Time-to-notification (Mean ± SD)	Not explicitly stated, but comparable to predicate	13.2 ± 2.9 seconds	25.8 ± 7.0 seconds

2. Sample Size Used for the Test Set and Data Provenance

ICH Test Set: 814 clinical anonymized cases
LVO Test Set: 476 clinical anonymized cases
Data Provenance: Retrospective, multinational study from 3 clinical sources (2 US and 1 OUS).
- For LVO positive cases: 156 (83%) were US and 32 (17%) OUS.
- The datasets contained a sufficient number of cases from important cohorts regarding imaging acquisitions (scanner makers: GE, Siemens, Philips, Toshiba/Canon; number of detector rows, gantry tilt, and slice thickness) and patient groups (age, sex, and US regions).

3. Number of Experts Used to Establish Ground Truth for the Test Set and Qualifications

Number of Experts: Three (3)
Qualifications of Experts: US-board-certified neuroradiologist readers. (Specific years of experience not mentioned).

4. Adjudication Method for the Test Set

Method: Concurrence of three US-board-certified neuroradiologist readers. (This implies a consensus or majority agreement method, often referred to as "3+0" or "majority rule" for establishing ground truth.)

5. Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study

The provided document does not mention a multi-reader multi-case (MRMC) comparative effectiveness study evaluating how human readers improve with or without AI assistance. The study described focuses on the standalone performance of the algorithm.

6. Standalone Performance Study

Yes, a standalone (algorithm only) performance study was done. The sections "IX. Performance Testing" and "IX.2. Performance Testing" explicitly describe the evaluation of the Cina software's performance (Sensitivity, Specificity, AUC, Accuracy, and Time-to-notification) directly against the established ground truth.

7. Type of Ground Truth Used

Type of Ground Truth: Expert consensus (established by the concurrence of three US-board-certified neuroradiologist readers, evaluating imaging findings). The text refers to "operators' visual assessments" which, in this context, refers to the expert readers' visual interpretation used to define the ground truth.

8. Sample Size for the Training Set

The document does not explicitly state the sample size for the training set. It mentions the algorithm uses "an artificial intelligence algorithm with database of images" but does not provide the size of this database or how it was used for training versus testing.

9. How Ground Truth for the Training Set Was Established

The document does not explicitly state how the ground truth for the training set was established. It only describes the method for the ground truth of the test set (concurrence of three neuroradiologists).

Summary

{0}------------------------------------------------

Image /page/0/Picture/0 description: The image shows the logo of the U.S. Food and Drug Administration (FDA). The logo consists of two parts: a symbol on the left and the FDA name on the right. The symbol on the left is a stylized image of a human figure, while the FDA name on the right is written in blue letters. The words "U.S. FOOD & DRUG ADMINISTRATION" are written in a clear, sans-serif font.

Avicenna.ai % John Smith Partner Hogan Lovells US LLP 555 13th Street, NW WASHINGTON, DISTRICT OF COLUMBIA 20004

November 22, 2022

Re: K221716

Trade/Device Name: Cina Regulation Number: 21 CFR 892.2080 Regulation Name: Radiological Computer Aided Triage And Notification Software Regulatory Class: Class II Product Code: OAS Dated: October 21, 2022 Received: October 25, 2022

Dear John Smith:

We have reviewed your Section 510(k) premarket notification of intent to market the device referenced above and have determined the device is substantially equivalent (for the indications for use stated in the enclosure) to legally marketed predicate devices marketed in interstate commerce prior to May 28, 1976, the enactment date of the Medical Device Amendments, or to devices that have been reclassified in accordance with the provisions of the Federal Food, Drug, and Cosmetic Act (Act) that do not require approval of a premarket approval application (PMA). You may, therefore, market the device, subject to the general controls provisions of the Act. Although this letter refers to your product as a device, please be aware that some cleared products may instead be combination products. The 510(k) Premarket Notification Database located at https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfpmn/pmn.cfm identifies combination product submissions. The general controls provisions of the Act include requirements for annual registration, listing of devices, good manufacturing practice, labeling, and prohibitions against misbranding and adulteration. Please note: CDRH does not evaluate information related to contract liability warranties. We remind you, however, that device labeling must be truthful and not misleading.

If your device is classified (see above) into either class II (Special Controls) or class III (PMA), it may be subject to additional controls. Existing major regulations affecting your device can be found in the Code of Federal Regulations, Title 21, Parts 800 to 898. In addition, FDA may publish further announcements concerning your device in the Federal Register.

Please be advised that FDA's issuance of a substantial equivalence determination does not mean that FDA has made a determination that your device complies with other requirements of the Act or any Federal

{1}------------------------------------------------

statutes and regulations administered by other Federal agencies. You must comply with all the Act's requirements, including, but not limited to: registration and listing (21 CFR Part 807); labeling (21 CFR Part 801); medical device reporting of medical device-related adverse events) (21 CFR 803) for devices or postmarketing safety reporting (21 CFR 4, Subpart B) for combination products (see https://www.fda.gov/combination-products/guidance-regulatory-information/postmarketing-safety-reportingcombination-products); good manufacturing practice requirements as set forth in the quality systems (QS) regulation (21 CFR Part 820) for devices or current good manufacturing practices (21 CFR 4, Subpart A) for combination products; and, if applicable, the electronic product radiation control provisions (Sections 531-542 of the Act); 21 CFR 1000-1050.

Also, please note the regulation entitled, "Misbranding by reference to premarket notification" (21 CFR Part 807.97). For questions regarding the reporting of adverse events under the MDR regulation (21 CFR Part 803), please go to https://www.fda.gov/medical-device-safety/medical-device-reportingmdr-how-report-medical-device-problems.

For comprehensive regulatory information about mediation-emitting products, including information about labeling regulations, please see Device Advice (https://www.fda.gov/medicaldevices/device-advice-comprehensive-regulatory-assistance) and CDRH Learn (https://www.fda.gov/training-and-continuing-education/cdrh-learn). Additionally, you may contact the Division of Industry and Consumer Education (DICE) to ask a question about a specific regulatory topic. See the DICE website (https://www.fda.gov/medical-device-advice-comprehensive-regulatoryassistance/contact-us-division-industry-and-consumer-education-dice) for more information or contact DICE by email (DICE@fda.hhs.gov) or phone (1-800-638-2041 or 301-796-7100).

Sincerely,

Jessica Lamb

Jessica Lamb, Assistant Director Imaging Software Team DHT8B: Division of Radiological Imaging Devices and Electronic Products OHT8: Office of Radiological Health Office of Product Evaluation and Quality Center for Devices and Radiological Health

Enclosure

{2}------------------------------------------------

DEPARTMENT OF HEALTH AND HUMAN SERVICES Food and Drug Administration Indications for Use

Form Approved: OMB No. 0910-0120 Expiration Date: 06/30/2023 See PRA Statement on last page

510(k) Number	K221716
Device Name	Cina

Indications for Use (Describe)

Cina is a radiological computer aided triage and notification software in the analysis of (1) not-enhanced head CT images and (2) CT angiography of the head.

Type of Use (Select one or both, as applicable)

Prescription Use (Part 21 CFR 801 Subpart D)

_ Over-The-Counter Use (21 CFR 801 Subpart C)

PLEASE DO NOT WRITE BELOW THIS LINE - CONTINUE ON A SEPARATE PAGE IF NEEDED.

FOR FDA USE ONLY

Concurrence of Center for Devices and Radiological Health (CDRH) (Signature)

This section applies only to requirements of the Paperwork Reduction Act of 1995.

DO NOT SEND YOUR COMPLETED FORM TO THE PRA STAFF EMAIL ADDRESS BELOW.

The burden time for this collection of information is estimated to average 79 hours per response, including the time to review instructions, search existing data sources, gather and maintain the data needed and complete and review the collection of information. Send comments regarding this burden estimate or any other aspect of this information collection, including suggestions for reducing this burden, to: Department of Health and Human

Services Food and Drug Administration

Office of Chief Information Officer

Paperwork Reduction Act (PRA)

Staff PRAStaff@fda.hhs.gov

"An agency may not conduct or sponsor, and a person is not required to respond to, a collection of information unless it displays a currently valid OMB number. "

FORM FDA 3881 (6/20)

{3}------------------------------------------------

510(k) SUMMARY

AVICENNA.AI's Cina

l. Submitter

Applicant:

AVICENNA.AI Espace Mistral- Batiment A 297 Avenue du Mistral 13600 La Ciotat France

Contact Person:

Stephane Berger Regulatory Manager Phone: +33 612122813 E-mail: stephane.berger@avicenna.ai

John J. Smith, MD, JD Partner, Hogan Lovells US LLP Phone: +1 202 637 3638 E-mail: john.smith@hoganlovells.com

Date prepared: November, 22, 2022

II. Device Identification

Name of Device:	Cina
Classification Name:	Radiological Computer-Assisted Triage AndNotification Software
Regulation No:	21 CFR § 892.2080
Product Code:	QAS
Regulatory Class:	Class II
Classification Panel:	Radiology devices

III. Predicate Device

The Cina is claimed to be substantially equivalent to Cina (K200855).

IV. Purpose of the Special 510(k) Notice

The modifications to Cina consist in minor software changes and labeling updates.

The main following modification that has been made is to the indications for use of the device within the labeling:

Indications for use have been modified with new information to include:

{4}------------------------------------------------

"large vessel occlusion (LVQ) of the anterior circulation (distal ICA, MCA-M1 or proximal MCA-M2)."

V. Intended Use / Indications for Use

Cina is a radiological computer aided triage and notification software indicated for use in the analysis of (1) non-enhanced head CT images and (2) CT angiography of the head. The device is intended to assist hospital networks and trained radiologists in workflow triage by flagging and communicating suspected positive findings of (1) head CT images for Intracranial Hemorrhage (ICH) and (2) head CT angiography for large vessel occlusion (LVO) of the anterior circulation (distal ICA, MCA-M1 or proximal MCA-M2).

Cina uses an artificial intelligence algorithm to analyze images and highlight cases with detected (1) ICH or (2) LVO on a standalone Web application in parallel to the ongoing standard of care image interpretation. The user is presented with notifications for cases with suspected ICH or LVO findings. Notifications include compressed preview images that are meant for informational purposes only, and are not intended for diagnostic use bevond notification. The device does not alter the original medical image, and it is not intended to be used as a diagnostic device.

VI. Device Description

Cina is a radiological computer-assisted triage and notification software device.

The software system is based on algorithm-programmed components and is comprised of a standard off-the-shelf operating system and additional image processing applications.

VII. Summary of Technological Characteristics

The subject and predicate devices have the same technological characteristics.

{5}------------------------------------------------

Cina runs on a standard "off the shelf" server/workstation and is made of the following software components:

. Cina Image Processing Applications including two applications: ICH and LVO;
. Cina Platform server that includes the Worklist client application in which notifications from the Cina Image Processing applications (ICH and LVO) are received.

Cina receives scans identified by the Cina Platform or other compatible medical image communications device, processes them using algorithmic methods involving execution of multiple computational steps to identify suspected presence of ICH or LVO and generates results files to be transferred by Cina Platform or a similar medical image communications device for output to a PACS system or workstation for worklist prioritization. Each of these components is briefly described below.

VII.1. Cina Platform

The Cina platform is an example of medical image communications platform for integrating and deploying the Cina ICH and LVO image processing applications. It provides the necessary requirements for interoperability based on the standardized DICOM protocol and services to communicate with existing systems in the hospital radiology department such as CT modalities or other DICOM nodes (DICOM router or PACS for example). It is responsible for transferring, storing, converting formats, notifying of suspected findings and displaying medical device data such as radiological data. The Cina Platform server includes the Worklist client application in which notifications from the Cina Image Processing applications (ICH and LVO) are received.

VII.2. ICH Application

The ICH application includes the software algorithm responsible for identifying and quantifying image characteristics that are consistent with an ICH. This application reads provided DICOM files, checks the DICOM properties to verify the compatibility with the recommended acquisition protocol. launches the algorithm and provides notification results (when an ICH is detected) compatible with the Cina Platform and with DICOM format.

VII.3. LVO Application

The LVO application includes the software algorithm responsible for identifying and quantifying image characteristics that are consistent with an LVO. This application reads provided DICOM files, checks the DICOM properties to verify the compatibility with recommended acquisition protocol, launches the algorithm part and provides notification results (when a LVO is detected) compatible with the Cina Platform and with DICOM format.

VIII. Substantial Equivalence

The subject and predicate devices have a similar intended use, technological characteristics, and principles of operation. The only difference is that the intended use of the subject has been revised to include the vessels (arteries) for which the device was designed and tested to detect LVO (the anterior circulation: distal ICA, MCA-M1 or proximal MCA-M2). Both devices are intended to provide the users with notifications and unannotated preview images of suspect studies for the purpose of preemptive triage, and are therefore substantially equivalent. A table comparing the key features of the subject and predicate devices is provided below.

{6}------------------------------------------------

Intended Use /Indications forUse	Subject device: Cina Software	Predicate device: Cina software(K200855)
	Cina is a radiological computeraided triage and notificationsoftware indicated for use in theanalysis of (1) non-enhanced headCT images and (2) CT angiographyof the head.The device is intended to assisthospital networks and trainedradiologists in workflow triage byflagging and communicatingsuspected positive findings of (1)head CT images for IntracranialHemorrhage (ICH) and (2) CTangiography of the head for largevessel occlusion (LVO) of theanterior circulation (distal ICA,MCA-M1 or proximal MCA-M2).	Cina is a radiological computer aidedtriage and notification softwareindicated for use in the analysis of (1)non-enhanced head CT images and(2) CT angiographies of the head.The device is intended to assisthospital networks and trainedradiologists in workflow triage byflagging and communicatingsuspected positive findings of (1) headCT images for IntracranialHemorrhage (ICH) and (2) CTangiographies of the head for largevessel occlusion (LVO).
	Cina uses an artificial intelligencealgorithm to analyze images andhighlight cases with detected (1)ICH or (2) LVO on a standaloneWeb application in parallel to theongoing standard of care imageinterpretation. The user ispresented with notifications forcases with suspected ICH or LVOfindings.	Cina uses an artificial intelligencealgorithm to analyze images andhighlight cases with detected (1) ICHor (2) LVO on a standalone Webapplication in parallel to the ongoingstandard of care image interpretation.The user is presented withnotifications for cases with suspectedICH or LVO findings.
	Notifications include compressedpreview images that are meant forinformational purposes only, andare not intended for diagnostic usebeyond notification. The devicedoes not alter the original medicalimage, and it is not intended to beused as a diagnostic device.	Notifications include compressedpreview images that are meant forinformational purposes only, and arenot intended for diagnostic use beyondnotification. The device does not alterthe original medical image, and it isnot intended to be used as adiagnostic device.
	The results of Cina are intended tobe used in conjunction with otherpatient information and based onprofessional judgment to assist withtriage/prioritization of medicalimages. Notified clinicians areultimately responsible for reviewingfull images per the standard of care.	The results of Cina are intended to beused in conjunction with other patientinformation and based on professionaljudgment to assist withtriage/prioritization of medical images.Notified clinicians are ultimatelyresponsible for reviewing full imagesper the standard of care.
	Subject device: Cina Software	Predicate device: Cina software (K200855)
Userpopulation	Radiologist	Radiologist
Anatomicalregion ofinterest	Head	Head
Dataacquisitionprotocol	Non contrast CT scan of the heador neck and CT angiogram imagesof the brain	Non contrast CT scan of the head orneck and CT angiogram images of thebrain
View DICOMdata	DICOM information about thepatient, study and current image	DICOM information about the patient,study and current image
Segmentationof region ofinterest	No; device does not mark, highlight,or direct users' attention to aspecific location in the originalimage	No; device does not mark, highlight, ordirect users' attention to a specificlocation in the original image
Algorithm	Artificial intelligence algorithm withdatabase of images	Artificial intelligence algorithm withdatabase of images
Notification /Prioritization	Yes	Yes
Previewimages	Presentation of a preview of thestudy for initial assessment notmeant for diagnostic purposes.The device operates in parallel withthe standard of care, which remainsthe default option for all cases.	Presentation of a preview of the studyfor initial assessment not meant fordiagnostic purposes.The device operates in parallel withthe standard of care, which remainsthe default option for all cases.
Alteration oforiginal image	No	No
Removal ofcases fromworklist queue	No	No
Structure	- LVO and ICH image processingapplications- Cina Platform (worklist and ImageViewer)	- LVO and ICH image processingapplications- Cina Platform (worklist and ImageViewer)

Table 1: Substantial Equivalence Chart

{7}------------------------------------------------

IX. Summary of Performance Data

The following performance data were provided in support of the substantial equivalence determination.

IX.1. Software Verification and Validation Testing

The Cina device has been evaluated and verified in accordance with software specifications and applicable performance standards through Software Development and Validation & Verification Process to ensure performance according to specifications, User Requirements and Federal Regulations and Guidance documents, "Guidance for the Content of Premarket Submissions for Software Contained in Medical Devices".

IX. 2. Performance Testing

{8}------------------------------------------------

Avicenna.Al conducted a retrospective, blinded, multinational study with the Cina software with the primary endpoint to evaluate the software's performance in 1) non-contrast CT (NCCT) head images pertaining to patient with suspected intracranial hemorrhage (ICH) findings and 2) CT angiography (CTA) head series pertaining to patient with suspected large vessel occlusion (LVO) findings, in 814 and 476 clinical anonymized cases, respectively. The device's Sensitivity and Specificity were analyzed, in addition to time-to-notification.

The data was provided from 3 clinical sources (2 US and 1 OUS). There were 255 (31.3%) positive ICH (images with ICH) and 188 (39.5%) positive LVO (images with LVO) cases included in the analysis. Within the 188 LVO positive cases, 156 (83%) were US and 32 (17%) OUS. Both tested dataset (for ICH and LVO) contained a sufficient number of cases from important cohorts in terms of imaging acquisitions (e.g., scanner makers – GE, Siemens, Philips and Toshiba/Canon; number of detector rows, gantry tilt and slice thickness) and patients' groups (e.g., age, sex and US regions).

Device sensitivities and specificities were compared to ground truth established by concurrence of three US-board-certified neuroradiologist readers.

Sensitivity and Specificity for the "ICH" prioritization and triage application are 91.4% (95% Cl: 87.2% – 94.5%) and 97.5% (95.8% – 98.6%), respectively. These findings achieved the 80% performance goal and are the same as those reported for Cina - ICH (K200855, the predicate device.

The ROC curve shows an AUC of 0.94, which is also the same as for the predicate device.

Regarding the "LVO" prioritization and triage application, Sensitivity and Specificity of 97.9% (95% Cl: 94.6% - 99.4%) and 97.6% (95% Cl: 95.1% - 99%), respectively are observed. These results achieved the 80% performance goal and are the same as the ones reported for Cina - LVO (K200855), the predicate device.

The ROC curve shows an AUC of 0.98, which is also the same as for the predicate device.

The results of the standalone assessment study demonstrated an overall agreement (accuracy) of 95.6% and 97.7% for the "ICH" and "LVO" tested cases, respectively, when compared to the ground truth (operators' visual assessments).

Positive predictive value (PPV) and negative predictive value (NPV) with varying prevalence, for both applications, are presented in Table 1 below:

Prevalence	Cina - ICH triage application		Cina - LVO triage application
	PPV (%)	NPV (%)	PPV (%)	NPV (%)
10%	80.2	99.0	81.7	99.8
15%	86.6	98.5	87.7	99.6
20%	90.1	97.8	91.0	99.5
25%	92.4	97.1	93.1	99.3
30%	94.0	96.3	94.5	99.1
35%	95.2	95.5	95.6	98.8

Table 1: PPV and NVP values for ICH and LVO image processing applications

{9}------------------------------------------------

40%	96.1	94.4	96.4	98.6
45%	96.8	93.2	97.1	98.2
50%	97.3	91.9	97.6	97.9

Additionally, both "ICH" and "LVO" prioritization and triage effectiveness were evaluated by the standalone per-case processing time of the device (time-to-notification), the results are presented in Table 2 below:

Table 2: Time-to-notification for ICH and LVO image processing applications

Time-to-Notification	MEAN ± SD(seconds)	MEDIAN(seconds)	Lower95% CI(seconds)	Upper95% CI(seconds)	MIN(seconds)	MAX(seconds)
Cina-ICH(N = 814)	$13.2 \pm 2.9$	13.2	13.0	13.4	8.6	39.1
Cina-LVO(N = 476)	$25.8 \pm 7.0$	24.7	25.1	26.4	13.0	55.3

The standalone effectiveness assessment demonstrated a substantial equivalence of the Cina -ICH triage application when compared to the predicate device (Cina-ICH - K200855). Specifically, the Cina's "ICH" triage mean ± SD "time-to-notification" is estimated to 13.2 ± 2.9 seconds. This is similar to the one reported for the predicate device (Cina-ICH - K200855: 21.6 ± 4.4 seconds).

Regarding Cina - LVO triage application, the mean ± SD "time-to-notification" is estimated to 25.8 ± 7.0 seconds. This demonstrated a substantial equivalence with the predicate device (Cina-LVO - K200855: 34.7 ± 10.7 seconds).

The performance testing of the Cina device establishes that the subject device is as safe and effective as the predicate device. This established that the Cina device meets its intended use statement and is substantially equivalent to the predicate device.

X. Conclusions

The subject Cina device is as safe and effective as the predicate Cina, with the similar intended use, technological characteristics, and principles of operation. Including in the intended use the vessels (arteries) for which the device was designed and tested to detect LVO does not raise new or different questions of safety or effectiveness.

Regulation Number and Section

§ 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.