(241 days)
VELMENI for DENTISTS (V4D) is a concurrent-read, computer-assisted detection software intended to assist dentists in the clinical detection of dental caries, fillings/restorations, fixed prostheses, and implants in digital bitewing, periapical, and panoramic radiographs of permanent teeth in patients 15 years of age or older. This device provides additional information for dentists in examining radiographs of patients' teeth. This device is not intended as a replacement for a complete examination by the dentist or their clinical judgment that considers other relevant information from the image, patient history, or actual in vivo clinical assessment. Final diagnoses and patient treatment plans are the responsibility of the dentist.
V4D software medical device comprises of the following key components:
- Web Application Interface delivers front-end capabilities and is the point of interaction between the device and the user.
- Machine Learning (ML) Engine delivers V4D's core ML capabilities through the radiograph type classifier, condition detection module, tooth numbering module, and merging module.
- Backend API allows interaction between all the components, as defined in this section, in order to fulfill the user's requests on the web application interface.
- Queue receives and stores messages from Backend API to send to Al-Worker.
- Al-Worker accepts radiograph analysis requests from Backend API via the Queue, passes gray scale radiographs to the ML Engine in the supported extensions (jpeg and png), and returns the ML analysis results to the Backend API.
- Database and File Storage store critical information related to the application, including user data, patient profiles, analysis results, radiographs, and associated data.
The following non-medical interfaces are also available with VELMENI for DENTISTS (V4D):
- VELMENI BRIDGE (VB) acts as a conduit enabling data and information exchange between Backend API and third-party software like Patient Management or Imaging Software
- Rejection Review (RR) module captures the ML-detected conditions rejected by dental professionals to aid in future product development and to be evaluated in accordance with VELMENIs post-market surveillance procedure.
Here's a breakdown of the acceptance criteria and the study proving the device meets them, based on the provided FDA 510(k) summary for "Velmeni for Dentists (V4D)":
1. Table of Acceptance Criteria and Reported Device Performance
The document does not explicitly present "acceptance criteria" in a tabular format with predefined thresholds. Instead, it reports the performance metrics from both standalone and clinical (MRMC) studies. The acceptance criteria are implicitly met if the reported performance demonstrates safety, effectiveness, and substantial equivalence to the predicate device.
Implicit Acceptance Criteria & Reported Device Performance:
| Metric / Feature | Acceptance Criteria (Implicit) | Reported Device Performance (Velmeni for Dentists (V4D)) |
|---|---|---|
| Standalone Performance | Demonstrate objective performance (sensitivity, specificity, Dice coefficient) for the indicated features across supported image types. | Caries (Lesion-Level Sensitivity): Bitewing: 72.8%, Periapical: 70.6%, Panoramic: 68.3% Fixed Prosthesis (Lesion-Level Sensitivity): Bitewing: 92.1%, Periapical: 81.0%, Panoramic: 74.5% Implant (Lesion-Level Sensitivity): Bitewing: 81.1%, Periapical: 94.5%, Panoramic: 79.6% Restoration (Lesion-Level Sensitivity): Bitewing: 88.1%, Periapical: 76.8%, Panoramic: 72.6% False Positives Per Image (Mean): Caries: 0.24-0.33, Fixed Prosthesis: 0.01-0.06, Implant: 0.00-0.01, Restoration: 0.10-0.62 Dice Score (Mean): Caries: 77.07-82.77%, Fixed Prosthesis: 91.47-97.09%, Implant: 88.67-95.47%, Restoration: 81.49-90.45% |
| Clinical Performance (MRMC) | Demonstrate that human readers (dentists) improve their diagnostic performance (e.g., sensitivity, wAFROC AUC) when assisted by the AI device, compared to working unassisted, without an unacceptable increase in false positives or decrease in specificity. The device should provide clear benefit. | wAFROC AUC (Aided vs. Unaided): Bitewing: 0.848 vs. 0.794 (Diff: 0.054), Periapical: 0.814 vs. 0.721 (Diff: 0.093), Panoramic: 0.615 vs. 0.579 (Diff: 0.036)Significant Improvements in Lesion-Level Sensitivity, Case-Level Sensitivity, and/or reductions in False Positives per Image (details in study section below). The study states "The V4D software demonstrated clear benefit for bitewing and periapical views in all features. The panoramic view demonstrated benefit..." |
| Safety & Effectiveness | The device must be demonstrated to be as safe and effective as the predicate device, with any differences in technological characteristics not raising new or different questions of safety or effectiveness. | "The results of the stand-alone and MRMC reader studies demonstrate that the performance of V4D is as safe, as effective, and performs equivalent to that of the predicate device, and VELMENI has demonstrated that the proposed device complies with applicable Special Controls for Medical Image Analyzers. Therefore, VELMENI for DENTISTS (V4D) can be found substantially equivalent to the predicate device." |
2. Sample Sizes Used for the Test Set and Data Provenance
-
Test Set Sample Sizes:
- Standalone Performance:
- 600 Bitewing images
- 597 Periapical images
- 600 Panoramic images
- Clinical Performance (MRMC):
- 600 Bitewing images (total caries 315)
- 597 Periapical images (total caries 271)
- 600 Panoramic images (total caries 853)
- Standalone Performance:
-
Data Provenance: The document states that "Subgroup analyses were performed among types of caries (primary and secondary caries; for caries-level sensitivity only), sex, age category, sensor, and study site." This suggests the data was collected from multiple study sites, implying a degree of diversity in the source of the images. However, the specific country of origin of the data is not explicitly stated. The study for initial data development seems to be centered around US licensed dentists and oral radiologists, suggesting US-based data collection, but this is not definitively stated for the entire dataset. The images were collected from various sensor manufacturers (Dexis, Dexis platinum, Kavo, Carestream, Planmeca). The document does not explicitly state whether the data was retrospective or prospective.
3. Number of Experts Used to Establish the Ground Truth for the Test Set and Their Qualifications
- Number of Experts:
- Ground truth for both standalone and clinical performance studies was established by three US licensed dentists.
- Non-consensus labels were adjudicated by one oral radiologist.
- Qualifications of Experts:
- US licensed dentists.
- Oral radiologist.
- No further details on their experience (e.g., years of experience) are provided in this summary.
4. Adjudication Method for the Test Set
- Adjudication Method: Consensus with Adjudication.
- Ground truth was initially established by "consensus labels of three US licensed dentists."
- "Non-consensus labels were adjudicated by an oral radiologist." This implies a "3+1" or similar method, where the three initial readers attempt to reach consensus, and any disagreements are resolved by a fourth, independent expert (the oral radiologist).
5. If a Multi Reader Multi Case (MRMC) Comparative Effectiveness Study was done, and its effect size
-
Yes, an MRMC comparative effectiveness study was done. It was described as a "multi-reader fully crossed reader improvement study."
-
Effect Size of Human Readers' Improvement with AI vs. Without AI Assistance:
The effect size is presented as the difference in various metrics (wAFROC AUC, lesion-level sensitivity, case-level sensitivity) between aided and unaided modes.-
wAFROC AUC:
- Bitewing: +0.054 (Aided 0.848 vs Unaided 0.794)
- Periapical: +0.093 (Aided 0.814 vs Unaided 0.721)
- Panoramic: +0.036 (Aided 0.615 vs Unaided 0.579)
-
Lesion-Level Sensitivity Improvement (Aided vs. Unaided):
- Bitewing:
- Caries: +12.8% (80.3% vs 67.5%)
- Fixed Prosthesis: +5.5% (95.7% vs 90.2%)
- Implant: +32.0% (93.2% vs 61.3%)
- Restoration: +16.7% (90.8% vs 74.1%)
- Periapical:
- Caries: +24.8% (73.4% vs 48.7%)
- Fixed Prosthesis: +11.1% (91.1% vs 80.0%)
- Implant: +16.4% (95.9% vs 79.5%)
- Restoration: +10.3% (90.6% vs 80.3%)
- Panoramic:
- Caries: +6.5% (27.2% vs 15.1%)
- Fixed Prosthesis: +8.2% (88.8% vs 80.5%)
- Implant: +8.7% (88.3% vs 79.6%)
- Restoration: +15.6% (73.0% vs 57.4%)
- Bitewing:
-
The study design also included measures of false positives per image (Mean FPs per Image) and case-level specificity to evaluate potential adverse effects of aid. While some specificities slightly decreased (e.g., Periapical Caries: -10.3%), the document generally concludes "The V4D software demonstrated clear benefit for bitewing and periapical views in all features. The panoramic view demonstrated benefit though the absolute benefit for caries sensitivity was smaller due to lower overall reader performance. In addition, for the panoramic view, there was a benefit in restoration sensitivity that was somewhat offset by a drop in image-level specificity."
-
6. If a Standalone (i.e., algorithm only without human-in-the-loop performance) was done
- Yes, a standalone performance evaluation was done. It was conducted against the established ground truth. The results are reported in Tables 2 and 3 and include lesion-level sensitivity, case-level sensitivity, false positives per image, case-level specificity, and Dice coefficient for segmentation.
7. The Type of Ground Truth Used
- The ground truth used for both standalone and clinical studies was based on expert consensus with adjudication. Specifically, it was established by "consensus labels of three US licensed dentists, and nonconsensus labels were adjudicated by an oral radiologist."
8. The Sample Size for the Training Set
- The document does not provide the sample size for the training set. It only describes the test set and validation processes.
9. How the Ground Truth for the Training Set Was Established
- The document does not explicitly describe how the ground truth for the training set was established. It focuses solely on the ground truth establishment for the test (evaluation) dataset. It's common for training data ground truth to be established through similar expert labeling processes, but this is not mentioned in the provided summary.
{0}------------------------------------------------
August 30, 2024
Image /page/0/Picture/1 description: The image contains the logo of the U.S. Food and Drug Administration (FDA). The FDA logo is composed of two parts: the Department of Health & Human Services logo on the left and the FDA acronym and full name on the right. The Department of Health & Human Services logo is a stylized depiction of an eagle. The FDA acronym is in a blue square, and the full name "U.S. Food & Drug Administration" is in blue text.
Velmeni Inc. °/o Mini Suri CEO 333 West Maude Avenue Suite 207 SUNNYVALE, CA 94085
Re: K240003
Trade/Device Name: Velmeni for Dentists (V4D) Regulation Number: 21 CFR 892.2070 Regulation Name: Medical Image Analyzer Regulatory Class: Class II Product Code: MYN Dated: August 1, 2024 Received: August 1, 2024
Dear Mini Suri:
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 (the 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 available 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.
Additional information about changes that may require a new premarket notification are provided in the FDA guidance documents entitled "Deciding When to Submit a 510(k) for a Change to an Existing Device" (https://www.fda.gov/media/99812/download) and "Deciding When to Submit a 510(k) for a Software Change to an Existing Device" (https://www.fda.gov/media/99785/download).
{1}------------------------------------------------
Your device is also subject to, among other requirements, the Quality System (QS) regulation (21 CFR Part 820), which includes, but is not limited to, 21 CFR 820.30, Design controls; 21 CFR 820.90, Nonconforming product; and 21 CFR 820.100, Corrective and preventive action. Please note that regardless of whether a change requires premarket review, the QS regulation requires device manufacturers to review and approve changes to device design and production (21 CFR 820.30 and 21 CFR 820.70) and document changes and approvals in the device master record (21 CFR 820.181).
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 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 Part 803) for devices or postmarketing safety reporting (21 CFR Part 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 Part 4, Subpart A) for combination products; and, if applicable, the electronic product radiation control provisions (Sections 531-542 of the Act); 21 CFR Parts 1000-1050.
Also, please note the regulation entitled, "Misbranding by reference to premarket notification" (21 CFR 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).
Sincerelv.
Lu Jiang
Lu Jiang, Ph.D. Assistant Director Diagnostic X-ray Systems Team DHT8B: Division of Radiologic Imaging Devices and Electronic Products OHT8: Office of Radiological Health Office of Product Evaluation and Quality Center for Devices and Radiological Health
{2}------------------------------------------------
Indications for Use
Submission Number (if known)
Device Name
VELMENI for DENTISTS (V4D)
Indications for Use (Describe)
VELMENI for DENTISTS (V4D) is a concurrent-read, computer-assisted detection software intended to assist dentists in the clinical detection of dental caries, fillings/restorations, fixed prostheses, and implants in digital bitewing, periapical, and panoramic radiographs of permanent teeth in patients 15 years of age or older. This device provides additional information for dentists in examining radiographs of patients' teeth. This device is not intended as a replacement for a complete examination by the dentist or their clinical judgment that considers other relevant information from the image, patient history, or actual in vivo clinical assessment. Final diagnoses and patient treatment plans are the responsibility of the dentist.
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)
CONTINUE ON A SEPARATE PAGE IF NEEDED.
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."
{3}------------------------------------------------
VELMENI for DENTISTS (V4D)
510(k) Summary
In accordance with 21 CFR 807.87(h) and 21 CFR 807.92, the following 510(k) Summary for VELMENI for DENTISTS (V4D) is provided:
Submitter Information
| Submitter: | Velmeni Inc.333 West Maude Avenue, STE 207Sunnyvale, CA 94085Phone: 201-289-3500 |
|---|---|
| ------------ | ---------------------------------------------------------------------------------------------- |
June 29, 2023 Date Prepared:
Contact Person:
Mini Suri, CEO Velmeni Inc. Phone: 201-289-3500 Email: mini@velmeni.com
Identification of the Device
| Trade Name: | VELMENI for DENTISTS (V4D) |
|---|---|
| Common Name: | Medical image analyzer |
| Classification Name | Medical image analyzer21CFR892.2070 |
| Product Code: | MYN |
| Device Class: | Class II |
| Predicate Device(s) | |
| Predicate Device(s): | Overjet Caries Assist (K222746) |
| Reference Device: | Second Opinion (K210365) |
| Reference Device: | Denti.Al Detect (K230144) |
Intended Use/ Indications for Use
VELMENI for DENTISTS (V4D) is a concurrent-read, computer-assisted detection software intended to assist dentists in the clinical detection of dental caries, fillings/restorations, fixed prostheses, and implants in digital bitewing, periapical, and panoramic radiographs of permanent teeth in patients 15 years of age or older. This device provides additional information for dentists in examining radiographs of patients' teeth. This device is not intended as a replacement for a complete examination by the dentist or their clinical judgment that considers other relevant information from the image, patient history, or actual in vivo clinical assessment. Final diagnoses and patient treatment plans are the responsibility of the dentist.
{4}------------------------------------------------
Device Description
V4D software medical device comprises of the following key components:
- Web Application Interface delivers front-end capabilities and is the point of interaction between the device and the user.
- Machine Learning (ML) Engine delivers V4D's core ML capabilities through the radiograph type classifier, condition detection module, tooth numbering module, and merging module.
- Backend API allows interaction between all the components, as defined in this section, in order to fulfill the user's requests on the web application interface.
- Queue receives and stores messages from Backend API to send to Al-Worker. ●
- Al-Worker accepts radiograph analysis requests from Backend API via the Queue, ● passes gray scale radiographs to the ML Engine in the supported extensions (jpeg and png), and returns the ML analysis results to the Backend API.
- Database and File Storage store critical information related to the application, including ● user data, patient profiles, analysis results, radiographs, and associated data.
The following non-medical interfaces are also available with VELMENI for DENTISTS (V4D):
- . VELMENI BRIDGE (VB) acts as a conduit enabling data and information exchange between Backend API and third-party software like Patient Management or Imaging Software
- Rejection Review (RR) module captures the ML-detected conditions rejected by dental ● professionals to aid in future product development and to be evaluated in accordance with VELMENIs post-market surveillance procedure.
Substantial Equivalence
The proposed VELMENI for DENTISTS (V4D) has similar indications for use to, and uses the same fundamental technology as, the legally marketed predicate device to which substantial equivalence is claimed, the Overiet Caries Assist (K2222746) device. Reference devices are included for comparison for detection features and image types.
| Specification | ProposedDevice:Velmeni ForDentists(V4D) | Predicate Device:Overjet CariesAssistK222746 | ReferenceDevice:Denti.AlDetectK230144 | ReferenceDevice:SecondOpinionK210365 |
|---|---|---|---|---|
| Manufacturer | Velmeni Inc. | Overjet Inc. | Denti.AlTechnology Inc. | Pearl Inc. |
| Classification | 892.2070 | 892.2070 | 892.2070 | 892.2070 |
| ProductCode | MYN | MYN | MYN | MYN |
| Specification | ProposedDevice:Velmeni ForDentists(V4D) | Predicate Device:Overjet CariesAssistK222746 | ReferenceDevice:Denti.AlDetectK230144 | ReferenceDevice:SecondOpinionK210365 |
| ImageModality | Radiograph | Radiograph | Radiograph | Radiograph |
| IntendedUse | To aid in clinicaldetection ofpathologic and/ornon-pathologicaldental features inradiographs ofpermanent teeth | To aid in clinicaldetection ofpathologicand/or non-pathologicaldental featuresin radiographs ofpermanent teeth | To aid in clinicaldetection ofpathologic and/ornon-pathologicaldental features inradiographs ofpermanent teeth | To aid in clinicaldetection ofpathologic and/ornon-pathologicaldental features inradiographs ofpermanent teeth |
| Indications | Velmeni forDentists (V4D) isa concurrent-read,computer-assisteddetection softwareintended to assistdentist in theclinical detectionof dental caries,fillings/restorations, fixedprostheses, andimplants in digitalbitewing,periapical, andpanoramicradiographs ofpermanent teethin patients 15years of age orolder. This deviceprovidesadditionalinformation fordentists inexaminingradiographs ofpatients' teeth.This device is notintended as areplacement for acompleteexamination bythe dentist or theirclinical judgmentthat considersother relevant | The OverjetCaries Assist(OCA) is aradiological,automated,concurrent read,computer-assisteddetectionsoftware intendedto aid in thedetection andsegmentation ofcaries onbitewing andperiapicalradiographs. Thedevice providesadditionalinformation forthe dentist to usein their diagnosisof tooth surfacesuspected ofbeing carious.The device is notintended as areplacement forcompletedentist's reviewor their clinicaljudgment thattakes intoaccount otherrelevantinformation fromthe image, | Denti.Al Detect isa Computer-AssistedDetection (CADe)software deviceintended to beused by dentalprofessionals,comprisingdentists anddental specialists,while readingextraoral andintraoral 2Ddentalradiographs. Thedevice aims toassist indetecting andhighlightinguncategorizedregions ofinterest (ROIs)within the teetharea, whichinclude cariesand periapicalradiolucency, asa second reader.The device isalso intended toaid in themeasurements ofmesial and distalbone levelsassociated witheach tooth. Thedevice is aimed | SecondOpinion® is acomputer aideddetection("CADe")software toidentify and markregions inrelation tosuspected dentalfindings whichinclude Caries,Discrepancy atthe margin of anexistingrestoration,Calculus,Periapicalradiolucency,Crown (metal,including zirconia& non-metal),Filling (metal &non- metal), Rootcanal, Bridge,and Implants. Itis designed to aiddental healthprofessionals toreview bitewingand periapicalradiographs ofpermanent teethin patients 12years of age orolder as a |
| Specification | ProposedDevice:VelmeniForDentists(V4D) | Predicate Device:Overjet CariesAssistK222746 | ReferenceDevice:Denti.AlDetectK230144 | ReferenceDevice:SecondOpinionK210365 |
| information fromthe image,patient history,or actual in vivoclinicalassessment.Final diagnosisand patienttreatment plansare theresponsibility ofthe dentist. | patient history,and actual in vivoclinicalassessment. | to be used withimages from thepatients of 22years age andolder withoutremainingprimarydentition. Thedevice is notintended toreplace acompleteclinician's reviewor clinicaljudgment thatconsiders otherrelevantinformation fromthe image orpatient history. | second reader. | |
| IntendedBody Part | Dental/teeth | Dental/teeth | Dental/teeth | Dental/teeth |
| End User | LicensedDentalProfessional | Dentist | DentalProfessional | Dental Clinicians |
| Patient Population | Patients requiringdental services,all sexes, at least15 years of age,and withpermanentdentition | Patients requiringdental services,all sexes, at least12 years of age,and withpermanentdentition | Patientsrequiring dentalservices, allsexes, at least22 years of ageor older | Patientsrequiring dentalservices, allsexes, at least12 years of ageor older |
| Prescriptio nor OTC | Prescription Use | Prescription Use | PrescriptionUse | Prescription Use |
| ReaderWorkflow | Concurrent Read | Concurrent Read | Second Read | Second Read |
| ImageSource | JPG, JPEG, PNGor DCM, DEX,and RVG | JPG, PNG, EOP,JIF, DICOM | JPEG, JPG,TIFF, TIF,PNG, BMP,DICOM | RVG, DICOM,JPEG, TIFF,PNG |
| CloudHostedSoftware | Yes | Yes | Yes | Yes |
| Data Input | Digitalintraoral filesof bitewingand periapicalradiographsand digitalextraoral filesof panoramicradiograph | Digital files ofbitewing andperiapicalradiographs | Intraoral (bitewingand periapical)Extraoral(panoramic) | Digital intraoralfiles of bitewingand periapicalradiographs |
| Model | MachineLearning | MachineLearning | Machine Learning | MachineLearning |
| Tooth Numbering | Yes | Yes | Unknown | Unknown |
| Detection | Caries,restorations,fixedprostheses,and implants | Caries | Caries andperiapicalradiolucency | Caries, margindiscrepancy-MD,calculus.periapicalradiolucency-PR,crown, bridges,implants, rootcanals, andfillings |
| Segmentation | Yes | Yes | Regions of Interest(ROIs) | No |
| Bounding Boxes | Yes | Unknown | Unknown | Yes |
Table 1. Comparison of the Proposed Device. Predicate Device and Reference Devices
{5}------------------------------------------------
{6}------------------------------------------------
{7}------------------------------------------------
Technological Characteristics
As shown in Table 1 above, the proposed VELMENI for DENTISTS (V4D) device and the predicate device have similar indications for use and uses the same fundamental technology (i.e., machine learning technology). Differences in technological characteristics include that each device uses its own proprietary algorithm to analyze digital radiographs. The proposed device's output is not limited to caries but includes restorations, fixed prostheses, and implants. Reference devices were identified because they demonstrate the Agency's familiarity with computer-assisted technology and the ability to identify and detect conditions beyond caries and in radiographs other than bitewing and periapical. The differences in technological characteristics do not raise new or different questions of safety or effectiveness.
The proposed device and the predicate device are technologically equivalent as follows:
- Both devices are designed to process digital bitewing and periapical radiographs.
- Both devices use neural network-based computer algorithms. ●
- Both devices mark detections within the user interface with a graphical overlay on the radiograph.
- . Both devices produce near-instantaneous detection results.
- Both devices are for prescription use. ●
- Both devices utilize cloud-hosted solutions. ●
{8}------------------------------------------------
The proposed device and the predicate device are technologically different as follows:
- The proposed device can detect other dental features (i.e., restorations, fixed prostheses, . and implants) in addition to dental caries. The ability to detect other dental features has been cleared in other devices within this same requlation and product code.
- The proposed device is capable of making detections in panoramic radiographs in addition to the bitewing and periapical radiographs. The ability to detect dental features in panoramic radiographs has been cleared in other devices within the same requlation and product code.
- The proposed device will be used on patient images from 15 years old or older with permanent teeth. The use of the device in patients as young as 12 years of age with permanent teeth has already been cleared within the predicate device.
- . Once suspected findings are detected in a radiograph, the proposed device overlays the radiograph with segmentation on the radiograph or alternatively a bounding box outlining the detected regions of suspected caries, restorations, fixed prosthesis, and implants, while predicate device does so with segmented polygons outlining the detected regions of suspected caries or fills suspected caries.
- . The proposed device utilizes a different list of compatibility of image types as compared to the predicate device.
Performance Data
Biocompatibilitv Testing:
There are no direct or indirect patient-contacting components of the proposed device. Therefore, patient contact information and biocompatibility testing are not applicable for this device.
Electrical Safety and Electromagnetic Compatibility (EMC):
The proposed device is a software-only device. It contains no electric components, generates no electrical emissions, and uses no electrical energy of any type. Therefore, electrical safety and EMC testing is not applicable for this device.
Software Verification and Validation Tests:
Software verification and validation testing were conducted, and documentation was provided as recommended by FDA's Guidance for Industry and FDA Staff, "Content of Premarket Submission for Device Software Functions." Verification of the software was conducted to ensure that the product works as designed. Validation was conducted to validate the design and the performance of the device to meet user needs and intended uses. VELMENI FOR DENTISTS (V4D) passed all verification and validation software tests. Overall, the VELMENI for DENTISTS (V4D) was found to be safe and effective for all intended users, uses, and use environments.
Animal Testing:
Animal studies were not necessary to establish the substantial equivalence of this device.
Bench Testing and Clinical Testing:
Velmeni Inc. conducted performance testing according to FDA's Guidance for Industry and Food and Drug Administration Staff, "Computer-Assisted Detection Devices Applied to Radiology Images and Radiological Device Data- Premarket Notification (510(k)) Submissions," as part of the development of VELMENI for DENTISTS (V4D). Performance testing included standalone testing and clinical reader evaluation.
{9}------------------------------------------------
Standalone Performance .
Standalone performance of the Velmeni for Dentists (V4D) was evaluated on a total of 600 Bitewing images, 597 Periapical images and 600 Panoramic images. Standalone performance was compared to ground truth established by consensus labels of three US licensed dentists, and nonconsensus labels were adjudicated by an oral radiologist.
Standalone testing included images form the following sensor manufacturers: Dexis, Dexis platinum, and Kavo used for bitewing and periapical image type and Kavo, Carestream, Planmeca sensor used for panoramic image type.
Standalone study assessed sensitivity, specificity and dice coefficient.
Table 2, Lesion-Level Sensitivity for Caries
| Caries | |||
|---|---|---|---|
| View | Lesion-Level Sensitivity( 95% CI1 ) | ||
| Bitewing Views | 72.8% (68.0%, 77.4%) | ||
| Periapical Views | 70.6% (64.1%, 76.6%) | ||
| Panoramic Views | 68.3% (64.9%, 71.8%) |
1 Two-sided bootstrap 95% Cl based on images with one or more caries and resampling of subjects.
2 Image classification for Sensitivity: TP = image with all lesions correctly identified. FN = image with at least one
lesion not identified.
Table 3, Lesion-level and case-level sensitivity for caries, fixed prothesis, implants and restorations and dice coefficient for lesion segmentation.
| Assessment | Results | ||
|---|---|---|---|
| Bitewing Views | Periapical Views | Panoramic Views | |
| Lesion-level Sensitivity ( 95% CI¹ ) | |||
| Fixed Prosthesis | 92.1% [90.0%, 94.1%] | 81.0% [76.6%, 85.4%] | 74.5% [71.5%, 77.3%] |
| Implant | 81.1% [67.6%, 92.1%] | 94.5% [90.0%, 98.2%] | 79.6% [69.7%, 87.4%] |
| Restoration | 88.1% [86.1%, 90.1%] | 76.8% [72.0%, 81.4%] | 72.6% [70.6%, 74.6%] |
| Case-level Sensitivity² ( 95% CI¹ ) | |||
| Caries | 59.5% [51.7%, 66.9%] | 57.5% [50.0%, 65.1%] | 45.0% [39.7%, 50.5%] |
| Fixed Prosthesis | 81.1% [75.7%, 86.3%] | 71.7% [62.5%, 79.9%] | 44.3% [38.0%, 50.6%] |
| Implant | 75.0% [59.1%, 88.9%] | 91.7% [84.8%, 97.2%] | 60.0% [45.0%, 74.4%] |
| Restoration | 71.6% [67.0%, 76.2%] | 68.1% [61.2%, 74.5%] | 27.2% [23.3%, 31.2%] |
{10}------------------------------------------------
| False Positives Per Image Mean (95% CI¹) | |||
|---|---|---|---|
| Caries | 0.24 [0.19, 0.29] | 0.27 [0.22, 0.32] | 0.33 [0.28, 0.38] |
| Fixed Prosthesis | 0.03 [0.01, 0.05] | 0.01 [0.00, 0.02] | 0.06 [0.04, 0.08] |
| Implant | 0.00 [0.00, 0.01] | 0.00 [NA, NA] | 0.01 [0.00, 0.02] |
| Restoration | 0.15 [0.11, 0.19] | 0.10 [0.07, 0.13] | 0.62 [0.54, 0.70] |
| Case-level Specificity³ ( 95% CI¹) | |||
| Caries | 88.0% [84.4%, 91.2%] | 84.7% [81.4%, 87.9%] | 96.8% [94.6%, 98.6%] |
| Fixed Prosthesis | 98.2% [96.4%, 99.7%] | 99.7% [99.2%, 100.0%] | 97.7%[96.0%, 99.2%] |
| Implant | 100.0%[NA, NA] | 100.0% [NA, NA] | 100.0% [NA, NA] |
| Restoration | 93.3% [89.1%, 96.9%] | 95.1%[92.8%, 97.2%] | 83.3% [75.0%, 90.6%] |
| DICE Score Mean (95% CI¹) | |||
| Caries | 81.96% (80.81%, 83.10%) | 82.77% (81.41%, 84.13%) | 77.07% (76.25%, 77.89%) |
| Fixed Prosthesis | 97.09% (96.84%, 97.33%) | 96.23% (95.78%, 96.69%) | 91.47% (91.24%, 91.71%) |
| Implant | 94.20% (92.44%, 95.97%) | 95.47% (94.60%, 96.34%) | 88.67% (87.22%, 90.11%) |
| Restoration | 90.45% (90.06%, 90.84%) | 81% (88.97%, 90.64%) | 81.49% (81.19%, 81.78%) |
1 Two-sided 95% Cl obtained by bootstrapping analysis of subjects.
2 Image classification for Sensitivity: TP = image with all lesions correctly identified. FN = image with at least one lesion not identified.
3 Image classification for Specificity: FP = image without a positive lesion that has at least one false positive. TN = image without a positive lesion that has no false positives.
Subgroup analyses were performed among types of caries (primary and secondary caries; for caries-level sensitivity only), sex, age category, sensor, and study site.
Clinical Performance ●
Velmeni evaluated the performance of Velmeni for Dentists (V4D) in a multi-reader fully crossed reader improvement study. 12 US licensed dentists were asked to evaluate 600 bitewing images (total caries 315), 597 periapical images (total caries 271) and 600 panoramic images (total caries 853). Ground truth was established by the consensus labels of three US licensed dentists, and nonconsensus labels were adjudicated by an oral radiologist. Half of the data set contained unannotated images, and the second half contained radiographs that had been processed through the V4D model. Radiographs were presented to readers in alternating groups. In Session 1, readers were asked to outline suspected caries, fixed prosthesis, implant, and restorations and to review predictions from the V4D model. Each reader was asked to provide a rating of 25 - 100 for their confidence in the annotation (25 for lowest confidence, up to 100 for highest confidence). A 4-week washout period was utilized to limit recollection bias. Following the washout, the readers were presented with the same data set but with alternate grouping. i.e., if a reader saw a radiograph in the
{11}------------------------------------------------
unpredicted state in session 1, they were presented with the same radiograph with V4D predictions in session 2, and vice versa. Results were compared against a consensus ground truth, and the sensitivity, specificity, and weighted alternative free response receiver operating characteristic (wAFROC) were evaluated to characterize the performance of the readers with (assisted) and without (unassisted) viewing the model annotations.
Findings were first summarized at the lesion level and then at the case (view) level. The V4D software demonstrated clear benefit for bitewing and periapical views in all features. The panoramic view demonstrated benefit though the absolute benefit for caries sensitivity was smaller due to lower overall reader performance. In addition, for the panoramic view, there was a benefit in restoration sensitivity that was somewhat offset by a drop in image-level specificity.
Table 4: wAFROC AUC at lesion- level for Caries Detection by Reader Mode
| View | Measure | Results | ||
|---|---|---|---|---|
| Aided | Unaided | Difference (95% CI) | ||
| Bitewing | wAFROC AUC | 0.848 | 0.794 | 0.054 (0.035, 0.073) |
| Periapical1 | wAFROC AUC | 0.814 | 0.721 | 0.093 (0.066, 0.120) |
| Panoramic | wAFROC AUC | 0.615 | 0.579 | 0.036 (0.022, 0.050) |
1 Excludes reader 1784 for Periapical View since the reader did not have all results for all aided assessments.
| Assessment | Measure | Results | |||
|---|---|---|---|---|---|
| Aided (95% CI) | Unaided (95% CI) | Difference (95% CI) | |||
| Bitewing | |||||
| Caries | Lesion-Level Sensitivity | 80.3% | 67.5% | 12.8% (9.9%, 15.9%) | |
| Mean FPs per Image | 0.18 | 0.18 | 0.00 (-0.03, 0.03) | ||
| Case-level Sensitivity | 69.5% | 51.5% | 18.1% (13.7%, 22.5%) | ||
| Case-level Specificity | 85.7% | 88.5% | -2.8% (-6.0%, 0.3%) | ||
| Fixed Prosthesis | Lesion-Level Sensitivity | 95.7% | 90.2% | 5.5% (3.6%, 7.6%) | |
| Mean FPs per Image | 0.04 | 0.13 | -0.08 (-0.11, -0.05) | ||
| Case-level Sensitivity | 89.6% | 82.7% | 6.8% (3.0%, 10.6%) | ||
| Case-level Specificity | 97.9% | 98.7% | -0.8% (-2.7%, 0.8%) | ||
| Implant | Lesion-Level Sensitivity | 93.2% | 61.3% | 32.0% (21.4%, 41.3%) | |
| Mean FPs per Image | 0.00 | 0.00 | 0.00 (-0.01, 0.00) | ||
| Case-level Sensitivity | 91.1% | 59.2% | 31.8% (21.2%, 42.9%) | ||
| Case-level Specificity | 100.0% | 99.9% | 0.1% (0.0%, 0.2%) | ||
| Restoration | Lesion-Level Sensitivity | 90.8% | 74.1% | 16.7% (14.3%, 19.2%) | |
| Mean FPs per Image | 0.15 | 0.29 | -0.14 (-0.18, -0.09) | ||
| Case-level Sensitivity | 77.8% | 52.7% | 25.1% (21.3%, 29.1%) | ||
| Case-level Specificity | 92.5% | 94.0% | -1.4% (-5.6%, 2.5%) | ||
| Assessment | Measure | Results | |||
| Aided (95% CI) | Unaided (95% CI) | Difference (95% CI) | |||
| Periapical1 | |||||
| Caries | Lesion-Level Sensitivity | 73.4% | 48.7% | 24.8% (19.8%, 29.8%) | |
| Mean FPs per Image | 0.19 | 0.08 | 0.11 (0.08, 0.14) | ||
| Case-level Sensitivity | 59.0% | 33.6% | 25.5% (19.6%, 31.3%) | ||
| Case-level Specificity | 84.2% | 94.5% | -10.3% (-13.0%, -7.6%) | ||
| Fixed Prosthesis | Lesion-Level Sensitivity | 91.1% | 80.0% | 11.1% (8.0%, 14.5%) | |
| Mean FPs per Image | 0.01 | 0.04 | -0.02 (-0.03, -0.01) | ||
| Case-level Sensitivity | 82.7% | 67.1% | 15.7% (10.6%, 20.9%) | ||
| Case-level Specificity | 99.7% | 99.5% | 0.2% (-0.2%, 0.6%) | ||
| Implant | Lesion-Level Sensitivity | 95.9% | 79.5% | 16.4% (12.7%, 20.3%) | |
| Mean FPs per Image | 0.00 | 0.01 | -0.01 (-0.01, 0.00) | ||
| Case-level Sensitivity | 93.8% | 77.5% | 16.3% (12.2%,21.0%) | ||
| Case-level Specificity | 99.9% | 100.0% | -0.1% (-0.4%, 0.0%) | ||
| Restoration | Lesion-Level Sensitivity | 90.6% | 80.3% | 10.3% (8.2%, 12.4%) | |
| Mean FPs per Image | 0.07 | 0.05 | -0.02 (0.00, 0.03) | ||
| Case-level Sensitivity | 83.9% | 69.4% | 14.5% (11.2%, 17.7%) | ||
| Case-level Specificity | 94.9% | 97.6% | -2.7% (-4.9%, -0.8%) | ||
| Panoramic | |||||
| Caries | Lesion-Level Sensitivity | 27.2% | 15.1% | 6.5% (4.5%, 8.6%) | |
| Mean FPs per Image | 0.21 | 0.30 | -0.09 (-0.13, -0.06) | ||
| Case-level Sensitivity | 11.5% | 8.5% | 3.0% (0.6%, 5.5%) | ||
| Case-level Specificity | 95.1% | 94.6% | 0.5% (-1.7%, 2.4%) | ||
| Fixed Prosthesis | Lesion-Level Sensitivity | 88.8% | 80.5% | 8.2% (6.1%, 10.3%) | |
| Mean FPs per Image | 0.07 | 0.18 | -0.10 (-0.13, -0.07) | ||
| Case-level Sensitivity | 70.5% | 67.0% | 3.5% (0.5%, 6.3%) | ||
| Case-level Specificity | 97.6% | 99.0% | -1.4% (-2.8%, -0.1%) | ||
| Implant | Lesion-Level Sensitivity | 88.3% | 79.6% | 8.7% (0.2%, 15.9%) | |
| Mean FPs per Image | 0.01 | 0.01 | 0.00 (-0.01, 0.01) | ||
| Case-level Sensitivity | 77.1% | 77.1% | 0.0% (-10.4%, 10.0%) | ||
| Case-level Specificity | 100.0% | 100.0% | 0.0% (NA, NA) | ||
| Restoration | Lesion-Level Sensitivity | 73.0% | 57.4% | 15.6% (14.3%, 16.9%) | |
| Mean FPs per Image | 0.73 | 1.02 | -0.29 (-0.37, -0.22) | ||
| Case-level Sensitivity | 26.7% | 19.6% | 7.1% (4.8%, 9.5%) | ||
| Case-level Specificity | 85.8% | 96.2% | -10.4% (-16.5%, -4.8%) |
{12}------------------------------------------------
1 Excludes reader 1784 for Periapical View since the reader did not have all results for all aided assessments.
Subgroup analysis was also performed for readers' experience, gender, age, imaging sensors, collection sites and primary and secondary caries.
{13}------------------------------------------------
Conclusion
Velmeni Inc. has the same intended use as the predicate device, specifically to detect pathological and/or non-pathological conditions on dental radiographs for use by trained dental professionals. While there are some small differences in technological characteristics between the proposed device, VELMENI for DENTISTS (V4D), and the predicate device, these differences do not raise different questions of safety and effectiveness. The results of the stand-alone and MRMC reader studies demonstrate that the performance of V4D is as safe, as effective, and performs equivalent to that of the predicate device, and VELMENI has demonstrated that the proposed device complies with applicable Special Controls for Medical Image Analyzers. Therefore, VELMENI for DENTISTS (V4D) can be found substantially equivalent to the predicate device.
§ 892.2070 Medical image analyzer.
(a)
Identification. Medical image analyzers, including computer-assisted/aided detection (CADe) devices for mammography breast cancer, ultrasound breast lesions, radiograph lung nodules, and radiograph dental caries detection, is a prescription device that is intended to identify, mark, highlight, or in any other manner direct the clinicians' attention to portions of a radiology image that may reveal abnormalities during interpretation of patient radiology images by the clinicians. This device incorporates pattern recognition and data analysis capabilities and operates on previously acquired medical images. This device is not intended to replace the review by a qualified radiologist, and is not intended to be used for triage, or to recommend diagnosis.(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 image analysis algorithms including a description of the algorithm inputs and outputs, each major component or block, and algorithm limitations.
(ii) A detailed description of pre-specified performance testing methods and dataset(s) used to assess whether the device will improve reader performance as intended and to characterize the standalone device performance. Performance testing includes one or more standalone tests, side-by-side comparisons, or a reader study, as applicable.
(iii) Results from performance testing that demonstrate that the device improves reader performance in the intended use population when used in accordance with the instructions for use. The performance assessment must be based on appropriate diagnostic accuracy measures (
e.g., receiver operator characteristic plot, sensitivity, specificity, predictive value, and diagnostic likelihood ratio). The test dataset must contain a sufficient number of cases from important cohorts (e.g., subsets defined by clinically relevant confounders, effect modifiers, concomitant diseases, and subsets defined by image acquisition characteristics) such that the performance estimates and confidence intervals of the device for these individual subsets can be characterized for the intended use population and imaging equipment.(iv) 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; and cybersecurity).(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 reading protocol.
(iii) A detailed description of the intended user and user training that addresses appropriate reading protocols for the device.
(iv) A detailed description of the device inputs and outputs.
(v) A detailed description of compatible imaging hardware and imaging protocols.
(vi) 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 or for certain subpopulations), as applicable.(vii) Device operating instructions.
(viii) A detailed summary of the performance testing, including: test methods, dataset characteristics, results, and a summary of sub-analyses on case distributions stratified by relevant confounders, such as lesion and organ characteristics, disease stages, and imaging equipment.