(119 days)
Yes
The intended use and device description explicitly state the use of "machine learning techniques" and a "frozen deep learning algorithm" for image analysis.
No
The device is intended to analyze fetal ultrasound images and clips to detect heart views and quality criteria, acting as a reading aid during acquisition and interpretation. It is not designed to provide or deliver therapy.
No
The device is described as a "concurrent reading aid" and is "intended to assist HCPs in the completeness assessment of the fetal heart ultrasound examination," rather than providing a diagnosis or clinical assessment itself. Its primary output is "Examination completeness," indicating which heart views and quality criteria are verified by the software, not a diagnostic finding.
Yes
The device is explicitly described as "a software" and its function is to analyze images received from an ultrasound machine via an HDMI connection. It does not include or require any specific hardware components beyond the standard computing device it is installed on and the connection to the ultrasound machine.
Based on the provided information, no, this device is not an IVD (In Vitro Diagnostic).
Here's why:
- IVD Definition: In Vitro Diagnostics are medical devices used to perform tests on samples taken from the human body (like blood, urine, tissue) to provide information about a person's health. They are used outside the body.
- FETOLY-HEART's Function: FETOLY-HEART analyzes ultrasound images and clips of the fetal heart. These are images generated by an external medical device (the ultrasound machine) and are not samples taken from the body.
- Intended Use: The intended use is to analyze images and act as a concurrent reading aid during the acquisition and interpretation of fetal ultrasound images. This is focused on image analysis and interpretation, not on testing biological samples.
While FETOLY-HEART is a medical device that uses AI and image processing, its function and intended use clearly fall outside the scope of In Vitro Diagnostics. It is a software intended to assist healthcare professionals in interpreting medical imaging data.
Yes
The letter explicitly states, "FDA's substantial equivalence determination also included the review and clearance of your Predetermined Change Control Plan (PCCP)." This language directly satisfies the condition for a "Yes" answer according to the Key Decision Rules.
Intended Use / Indications for Use
Fetoly-Heart is intended to analyze fetal ultrasound image sequences using machine learning techniques to automatically detect heart views and quality criteria within the views. The device is intended for use as a concurrent reading aid during the acquisition and interpretation of fetal ultrasound images.
Fetoly-Heart is indicated for use during routine fetal heart examination of 2nd and 3rd trimester pregnancy (gestational age: from 17 to 40 weeks).
Product codes (comma separated list FDA assigned to the subject device)
IYN, IYO, QIH
Device Description
FETOLY-HEART is a software that aims at helping sonographers, obstetricians, radiologists, maternal-fetal medicine specialists, and pediatric cardiologists (designated as healthcare professionals i.e. HCPs) to perform fetal ultrasound examinations of the fetal heart in real-time. FETOLY-HEART can be used by HCPs during fetal ultrasound examinations in the second and third trimesters (gestational age window: from 17 to 40 weeks). The software is intended to assist HCPs in the completeness assessment of the fetal heart ultrasound examination in accordance with national and international guidelines.
To utilize FETOLY-HEART, the software needs to be installed on a hardware device which is connected to an Ultrasound Machine through an HDMI connection. The software receives ultrasound images captured by the connected Ultrasound Machine in real-time. The software's frozen deep learning algorithm, which was trained by supervised learning, analyzes images of this ultrasound image stream to detect heart views and quality criteria within those views. The software provides the following user-accessible information:
- . Examination completeness: the software displays in real-time which heart views and quality criteria are verified by the software during the examination. It is the main and principal output of the FETOLY-HEART device. The verified heart views and quality criteria are accessible by clinicians at any moment of the ultrasound examination, in real-time.
- . Completeness illustration: the software selects an image subset that illustrates the verified views and quality criteria. These images can be reviewed by clinicians to verify the views and criteria's presence. This is a secondary output of the FETOLY-HEART device. Optionally, clinicians can display detected quality criteria localization on selected images.
Mentions image processing
Yes
Mentions AI, DNN, or ML
Yes
Input Imaging Modality
Ultrasound
Anatomical Site
Fetal heart
Indicated Patient Age Range
Gestational age: from 17 to 40 weeks (2nd and 3rd trimester pregnancy)
Intended User / Care Setting
Sonographers, obstetricians, radiologists, maternal-fetal medicine specialists, and pediatric cardiologists (healthcare professionals) during routine fetal heart examination.
Description of the training set, sample size, data source, and annotation protocol
Not Found. The document states "This testing dataset originated from distinct clinical sites from which the data used during model development (training/validation) was sourced, ensuring testing independence." but does not provide details on the training/validation data.
Description of the test set, sample size, data source, and annotation protocol
Description of the test set: Dataset of 2,288 fetal ultrasound images across 480 patient cases, including full examination still images, cardiac clip frames and full examination video frames.
Sample size: 2,288 images from 480 patient cases.
Data source: 7 clinical sites in the United States. Cases are representative of the intended use population. Patient cases were retrospectively collected in reverse chronological order until at least 20 patient files per subgroup and an overall of 275 patient files were reached. One image maximum per view per patient case was selected by categorizing 12,934 images from the patient cases into heart views and randomly picking one image maximum per view, resulting in a total of 2,288 images.
Annotation protocol:
For view classification: A 2+1 ground truth procedure was used. Six annotators (3 sonographers and 3 OB/GYN doctors) were paired and assigned uniformly distributed batches of images. Each image was annotated by a pair of annotators as belonging to one of 6 views. Images with annotator agreement were considered ground truth. Images in which the pair of annotators disagreed were reviewed by an adjudicator, who made the final decision.
For quality criteria classification and localization: Each image was annotated by a pair of annotators who drew bounding boxes on present criteria. If their boxes had at least 50% overlap, their coordinates were averaged to form the ground truth. If the overlap was lower or there was a disagreement on the criterion presence, an adjudicator reviewed the boxes. The final decision regarding the presence was based on majority consensus among the adjudicator and annotators. The final decision for the criteria localization was based on the adjudicator's decision to either keep one of the annotator's boxes or draw a new one.
Summary of Performance Studies (study type, sample size, AUC, MRMC, standalone performance, key results)
Study type: Standalone performance testing.
Sample size: 2,288 fetal ultrasound images across 480 patient cases.
AUC: Not mentioned.
MRMC: Not mentioned.
Standalone performance:
- Fetal heart view detection: Sensitivity >= 85% (acceptance criterion) and specificity >= 85% (acceptance criterion).
- Quality criteria within heart views detection: Sensitivity >= 90% (acceptance criterion) and specificity >= 90% (acceptance criterion).
- Localization of bounding boxes of quality criteria: Mean Intersection over Union (IoU) of >= 50% (acceptance criterion).
Key results: - All acceptance criteria were met.
- The lowest 95% CI for sensitivity for all fetal heart views was 0.960 for Abdomen view. The lowest 95% CI for specificity for all fetal heart views was 0.977 for "Other view".
- The lowest 95% CI for sensitivity for all individual quality criteria was 0.860 (Right pulmonary vein). The lowest 95% CI for specificity for all individual quality criteria was 0.985 (Trachea / bronchi).
- The lowest 95% CI for mean IoU for all individual quality criteria was 0.491 (Trachea / bronchi).
- Performance metrics were analyzed for each subgroup (maternal age, gestational age, territory, BMI, scanner manufacturer, heart abnormality, races and ethnicities, clinical sites) and confounder (image digital quality, image type) to validate the model's robustness and generalizability.
Key Metrics (Sensitivity, Specificity, PPV, NPV, etc.)
Fetal heart view detection:
- Overall Sensitivity: Lowest 95% CI for sensitivity among all views was 0.960 (Abdomen view). Point estimates range from 0.976 to 0.987.
- Overall Specificity: Lowest 95% CI for specificity among all views was 0.977 ("Other view"). Point estimates range from 0.983 to 1.00.
Quality criterion detection:
- Overall Sensitivity: Lowest 95% CI for sensitivity among all quality criteria was 0.860 (Right pulmonary vein). Point estimates range from 0.903 to 0.990.
- Overall Specificity: Lowest 95% CI for specificity among all quality criteria was 0.985 (Trachea / bronchi). Point estimates range from 0.990 to 1.00.
- Mean Intersection over Union (mIoU): The lowest 95% CI for mIoU among all quality criteria was 0.491 (Trachea / bronchi). Point estimates range from 0.512 to 0.792.
Predicate Device(s): If the device was cleared using the 510(k) pathway, identify the Predicate Device(s) K/DEN number used to claim substantial equivalence and list them here in a comma separated list exactly as they appear in the text. List the primary predicate first in the list.
Reference Device(s): Identify the Reference Device(s) K/DEN number and list them here in a comma separated list exactly as they appear in the text.
Predetermined Change Control Plan (PCCP) - All Relevant Information for the subject device only (e.g. presence / absence, what scope was granted / cleared under the PCCP, any restrictions, etc).
Presence / Absence: Included.
Scope granted / cleared under the PCCP: PCCP is for machine learning (ML) algorithm modifications, specifically:
- Modification of training and/or validation datasets: Increase or recovery (in case of data drift) of FETOLY-HEART's performance.
- Modification of model training hyperparameters: Improvement and optimization of FETOLY-HEART's performance.
- Heart quality criteria addition/removal: Maintain alignment with quality criteria recommended for fetal heart screening in state-of-the-art international guidelines.
Restrictions:
- All algorithm modifications will be adequately trained, tuned, and locked prior to release of the software with the modified ML model.
- The PCCP does not include the implementation of adaptive algorithms that will continuously learn in the field.
- Implemented modifications to the FETOLY-HEART algorithm will be communicated to users via the software update notifications and through updated labelling.
- Modified models will be tested for superiority on the performance study test dataset which will contain new unseen data (for dataset and hyperparameter modifications).
- New quality criteria will be controlled using the same acceptance criteria as defined by secondary endpoints, with proper performance testing to ensure no decrease in test performance (for quality criteria addition/removal). This change only pertains to quality criteria belonging to one of the 5 heart views already included in FETOLY-HEART.
§ 892.1550 Ultrasonic pulsed doppler imaging system.
(a)
Identification. An ultrasonic pulsed doppler imaging system is a device that combines the features of continuous wave doppler-effect technology with pulsed-echo effect technology and is intended to determine stationary body tissue characteristics, such as depth or location of tissue interfaces or dynamic tissue characteristics such as velocity of blood or tissue motion. This generic type of device may include signal analysis and display equipment, patient and equipment supports, component parts, and accessories.(b)
Classification. Class II.
0
September 11, 2024
Image /page/0/Picture/1 description: The image shows the logo of the U.S. Food and Drug Administration (FDA). The logo consists of two parts: the Department of Health & Human Services logo on the left and the FDA text logo on the right. The FDA text logo is in blue and reads "FDA U.S. FOOD & DRUG ADMINISTRATION".
Diagnoly % Nima Akhlaghi Associate Director, Digital Health Regulatory Affairs MCRA, LLC 505 Park Avenue, 14th Floor New York, NY 10022
Re: K241380
Trade/Device Name: FETOLY-HEART Regulation Number: 21 CFR 892.1550 Regulation Name: Ultrasonic Pulsed Doppler Imaging System Regulatory Class: Class II Product Code: IYN, IYO, QIH Dated: August 9, 2024 Received: August 9, 2024
Dear Nima Akhlaghi:
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.
FDA's substantial equivalence determination also included the review and clearance of your Predetermined Change Control Plan (PCCP). Under section 515C(b)(1) of the Act, a new premarket notification is not required for a change to a device cleared under section 510(k) of the Act, if such change is consistent with an
1
established PCCP granted pursuant to section 515C(b)(2) of the Act. Under 21 CFR 807.81(a)(3), a new premarket notification is required if there is a major change or modification in the intended use of a device, or if there is a change or modification in a device that could significantly affect the safety or effectiveness of the device, e.g., a significant change or modification in design, material, chemical composition, energy source, or manufacturing process. Accordingly, if deviations from the established PCCP result in a major change or modification in the intended use of the device, or result in a change or modification in the device that could significantly affect the safety or effectiveness of the a new premarket notification would be required consistent with section 515C(b)(1) of the Act and 21 CFR 807.81(a)(3). Failure to submit such a premarket submission would constitute adulteration and misbranding under sections 501(f)(1)(B) and 502(o) of the Act, respectively.
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).
Your device is also subject to, among other requirements, the Quality System (OS) 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.
All medical devices, including Class I and unclassified devices and combination product device constituent parts are required to be in compliance with the final Unique Device Identification System rule ("UDI Rele"). The UDI Rule requires, among other things, that a device bear a unique device identifier (UDI) on its label and package (21 CFR 801.20(a)) unless an exception or alternative applies (21 CFR 801.20(b)) and that the dates on the device label be formatted in accordance with 21 CFR 801.18. The UDI Rule (21 CFR 830.300(a) and 830.320(b)) also requires that certain information be submitted to the Global Unique Device Identification Database (GUDID) (21 CFR Part 830 Subpart E). For additional information on these requirements, please see the UDI System webpage at https://www.fda.gov/medical-device-advicecomprehensive-regulatory-assistance/unique-device-identification-system-udi-system.
2
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 medical devices and radiation-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,
Yanna S. Kang -S
Yanna Kang, Ph.D. Assistant Director Mammography and Ultrasound Team DHT8C: Division of Radiological Imaging and Radiation Therapy Devices OHT8: Office of Radiological Health Office of Product Evaluation and Quality Center for Devices and Radiological Health
Enclosure
3
Indications for Use
Submission Number (if known)
Device Name
FETOLY-HEART
Indications for Use (Describe)
Fetoly-Heart is intended to analyze fetal ultrasound image sequences using machine learning techniques to automatically detect heart views and quality criteria within the views. The device is intended for use as a concurrent reading aid during the acquisition and interpretation of fetal ultrasound images.
Fetoly-Heart is indicated for use during routine fetal heart examination of 2nd and 3rd trimester pregnancy (gestational age: from 17 to 40 weeks).
Type of Use (Select one or both, as applicable)
X 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."
4
Image /page/4/Picture/0 description: The image shows a logo for a company called DIAGNOLY. The logo features a blue magnifying glass with a stack of three horizontal lines inside the lens. The word "DIAGNOLY" is written in blue, sans-serif font below the magnifying glass. The background of the image is white.
In accordance with 21 CFR 807.92 the 510(k) summary for FETOLY-HEART is provided below.
510(k) owner 1
| Owner | Diagnoly
60 Avenue Rockefeller
69008 Lyon, France
+33(0)4.78.76.85.75 |
|--------------------------|-----------------------------------------------------------------------------------------------------------------------------------------|
| Primary contact person | Ivan Voznyuk
Chief Executive Officer
Diagnoly
Phone: +33(0)6.95.87.04.55
Email: ivan@diagnoly.com |
| Secondary contact person | Nima Akhlaghi
Associate Director, Digital Health Regulatory Affairs
MCRA, LLC
Phone: 202.742.3889
Email: nakhlaghi@mcra.com |
| Date prepared | 2024-08-30 |
2 Device
| Trade Name | FETOLY-HEART |
|---|---|
| Classification name | Accessory to Ultrasonic Pulsed Doppler Imaging System, 21 CFR 892.1550 |
| Accessory to Ultrasonic Pulsed Echo Imaging System, 21 CFR 892.1560 | |
| Medical image management and processing system, 21 CFR 892.2050 | |
| Class | II |
| Product code | IYN (Primary) |
| IYO, QIH (secondary) |
Predicate device identification ന
The predicate device used for FETOLY-HEART is the cardiac-related component of Sonio Detect (K240406).
Additionally, a reference device was chosen for FETOLY-HEART based on its substantially equivalent technical characteristics of automatic extraction of views from a sequence of images. This reference
5
Image /page/5/Picture/0 description: The image shows a logo for a company called DIAGNOLY. The logo features a blue magnifying glass with a stylized image of a stack of objects inside the lens. The text "DIAGNOLY" is written in blue below the magnifying glass.
device corresponds to the SonoLystLive view suggestion feature within the Voluson Expert Series 22/20/18 (K220358).
Device description 4
FETOLY-HEART is a software that aims at helping sonographers, obstetricians, radiologists, maternal-fetal medicine specialists, and pediatric cardiologists (designated as healthcare professionals i.e. HCPs) to perform fetal ultrasound examinations of the fetal heart in real-time. FETOLY-HEART can be used by HCPs during fetal ultrasound examinations in the second and third trimesters (gestational age window: from 17 to 40 weeks). The software is intended to assist HCPs in the completeness assessment of the fetal heart ultrasound examination in accordance with national and international guidelines.
To utilize FETOLY-HEART, the software needs to be installed on a hardware device which is connected to an Ultrasound Machine through an HDMI connection. The software receives ultrasound images captured by the connected Ultrasound Machine in real-time. The software's frozen deep learning algorithm, which was trained by supervised learning, analyzes images of this ultrasound image stream to detect heart views and quality criteria within those views. The software provides the following user-accessible information:
- . Examination completeness: the software displays in real-time which heart views and quality criteria are verified by the software during the examination. It is the main and principal output of the FETOLY-HEART device. The verified heart views and quality criteria are accessible by clinicians at any moment of the ultrasound examination, in real-time.
- . Completeness illustration: the software selects an image subset that illustrates the verified views and quality criteria. These images can be reviewed by clinicians to verify the views and criteria's presence. This is a secondary output of the FETOLY-HEART device. Optionally, clinicians can display detected quality criteria localization on selected images.
| Heart views | Quality criteria within the views | |
|---|---|---|
| Image: Heart view | ||
| (A) ABD | ||
| Abdomen view | ||
| (n = 8) | (A1) Sp Spine | |
| (A2) IRb Left rib | ||
| (A3) rRb Right rib | ||
| (A4) Ao Descending aorta | ||
| (A5) VC Inferior vena cava | ||
| (A6) St Stomach | ||
| (A7) Uv Umbilical vein | ||
| (A8) Ap Thorax apex | ||
| (B1) Sp Spine | ||
| (B2) IRb | Left rib | |
| (B3) rRb | Right rib | |
| (B4) Ao | Descending aorta | |
| (B5) IPV | Left pulmonary vein | |
| (B6) rPV | Right pulmonary vein | |
| (B7) LA | Left atrium | |
| (B8) RA | Right atrium | |
| (B) 4CH | (B9) FOP | Foramen Ovale flap (Vieussens valve) |
| Four chamber view | ||
| (n = 19) | (B10) FO | Open Foramen Ovale |
| (B11) MV | Mitral valve | |
| (B12) TV | Tricuspid valve | |
| (B13) bCr | Connection between crux and atrial septum (vestibular septum) | |
| (B14) Cr | Atrioventricular valve offset in crux | |
| (B15) tCr | Connection between interventricular septum and crux | |
| (B16) IVS | Interventricular septum | |
| (B17) LV | Left ventricle | |
| (B18) RV | Right ventricle | |
| (B19) Str | Sternum | |
| (C1) LA | Left atrium | |
| (C2) aAo | Proximal ascending aorta | |
| (C) LVOT | (C3) SV | Semilunar valves |
| (C4) LV | Left ventricle | |
| (C5) IVS | Interventricular septum | |
| Left Ventricular Outflow Tract view | ||
| (n = 6) | (C6) RV | Right ventricle |
| (D1) dAo | Descending aorta | |
| (D2) Tr | Trachea / bronchi | |
| (D3) IPA | Left pulmonary artery | |
| (D4) Du | Ductus arteriosus | |
| (D5) rPA | Right pulmonary artery | |
| (D6) Or | Origin of pulmonary arteries | |
| (D) RVOT | (D7) S | Septum between pulmonary artery trunk and ascending aorta |
| Right Ventricular Outflow Tract view | ||
| (n = 10) | (D8) aAo | Ascending aorta |
| (D9) SVC | Superior vena cava | |
| (D10) PA | Pulmonary trunk | |
| (E1) Sp | Spine | |
| (E2) Tr | Trachea | |
| (E3) ES | Side space on the left of ductus / pulmonary artery | |
| (E4) PA | Main pulmonary artery | |
| (E5) Du | Ductus (Ductal arch) | |
| (E6) aAo | Ascending aorta | |
| (E) 3VX | ||
| (E7) aAr | Aortic arch | |
| Three vessels view | ||
| (n = 9) | (E8) SVC | Superior vena cava |
| (E9) Th | Thymus / sternum |
Definition of a complete examination 4.1
6
Image /page/6/Picture/0 description: The image shows a logo for a company called DIAGNOLY. The logo features a blue magnifying glass with a stylized image of a stack of three cones inside. The company name, DIAGNOLY, is written in blue text below the magnifying glass.
7
Image /page/7/Picture/0 description: The image shows a logo for a company called "Diagnoly". The logo features a blue magnifying glass with a stack of three cone-shaped objects inside the lens. The word "Diagnoly" is written in blue, sans-serif font below the magnifying glass.
Table 1. List of 52 quality criteria defining the 5 views recommended by the International Society of Ultrasound in Obstetrics and Gynecology for the foetal heart screening of 2nd and 3rd trimesters of pregnancy. Abbreviations from this table are corresponding to the abbreviations used within the software.
International and national guidelines- recommend 5 foetal cardiac views for routine ultrasound examination of 2nd and 3rd trimesters: (A) Abdomen view, (B) Four chamber view, (C) Left Ventricular Outflow Tract view, (D) Right Ventricular Outflow Tract view, (E) Three vessels view. The quality of these 5 heart views depends on the presence of 52 anatomical quality criteria within the views (Table 1). Thus, an examination can be defined as complete when all 5 heart views and their quality criteria are obtained by the HCP.
Functionality 1: completeness overview 4.2
The software assesses the completeness of the foetal ultrasound examination. It verifies whether all the information corresponding to the recommended guidelines for the foetal heart examination has been acquired. This information corresponds to the presence of 5 main foetal cardiac views and 52 quality criteria, detailed in the above section, allowing for compliance in cardiac screening.
Functionality 2: completeness illustration 4.3
This functionality was developed to enhance the security of the completeness evaluation which is done by the first module. It enables clinicians to verify the examination completeness overview, i.e. verified views and quality criteria by the software through the gallery page of the software interface. Operating continuously, it evaluates each image processed by the first module in real-time, retaining an image set as visual evidence of the verified heart views and quality criteria up to that moment.
FETOLY-HEART does not aim to select the 'best' or of 'high' diagnostic quality according to a given qualitative scale. Rather, FETOLY-HEART aims to select images illustrating the examination quantitative completeness in terms of verified views and quality criteria.
Indications for use ഗ
FETOLY-HEART is intended to analyse fetal ultrasound images and clips using machine learning techniques to automatically detect heart views and quality criteria within the views. The device is intended for use as a concurrent reading aid during the acquisition and interpretation of fetal ultrasound images.
FETOLY-HEART is indicated for use during routine fetal heart examination of 2nd and 3rd trimester pregnancy (gestational age: from 17 to 40 weeks).
1 Carvalho JS, Axt-Fliedner R, Chaoui R, Copel JA, Cuneo BF, Goff D, Gordin Kopylov L, Hecher K, Lee W, Moon-Grady A, Mousa HA, Munoz H, Paladini D, Prefumo F, Quarello E, Rychik J, Tutschek B, Wiechec M, Yagel S. ISUOG Practice Guidelines (updated): fetal cardiac screening. Ultrasound Obstet Gynecol 2023; 61: 788-803.
8
Image /page/8/Picture/0 description: The image shows the text "510(k) Summary - K241380". The text is written in a bold, sans-serif font. The words "510(k) Summary" are followed by a dash and then the code "K241380". The text is centered and appears to be the title or heading of a document.
Image /page/8/Picture/1 description: The image shows a logo for a company called DIAGNOLY. The logo features a blue magnifying glass with a stack of three objects inside. The word "DIAGNOLY" is written in blue letters below the magnifying glass.
6 Summary of the technological characteristics of FETOLY-HEART compared to the predicate device
| Aspect | Predicate device:
Sonio Detect
(Cardiac Features)
K240406 | Reference device:
Voluson Expert
18/20/22 (Cardiac-
related SonolystLive
Feature)
K220358 | Proposed device:
FETOLY-HEART | Comparison between
Proposed and
Predicate device |
|---------------------------------------------|----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| General | | | | |
| Manufacturer
name | Sonio | GE Healthcare | Diagnoly | NA |
| Device name | Cardiac-related
features in Sonio
Detect | Cardiac-related
SonolystLive in the
Voluson Expert
18/20/22 | FETOLY-HEART | NA |
| Product
code(s) | IYN (Primary)
IYO, QIH (secondary) | IYN (Primary)
IYO, ITX (Secondary) | IYN (Primary)
IYO, QIH (secondary) | Substantially
equivalent
Primary codes are the
same for all devices |
| Regulation
number | - Accessory to
Ultrasonic Pulsed
Doppler Imaging
System, 21 CFR
892.1550
- Accessory to
Ultrasonic Pulsed
Echo Imaging System,
21 CFR 892.1560 | - Ultrasonic Pulsed
Doppler Imaging
System, 21 CFR
892.1550 - Ultrasonic Pulsed
Echo Imaging System,
21 CFR 892.1560 - Diagnostic
Ultrasound
Transducer, 21 CFR
892.1570, 90-ITX | - Accessory to
Ultrasonic Pulsed
Doppler Imaging
System, 21 CFR
892.1550 - Accessory to
Ultrasonic Pulsed
Echo Imaging System,
21 CFR 892.1560 - Medical image
management and
processing system,
21 CFR 892.2050 | Substantially
equivalent
All devices are class II
devices subject to
510(k) regulatory
pathway. |
| Brief
description | The predicate device
is a software that
aims at helping
sonographers,
OB/GYNs, MFMs and
Fetal surgeons (all
three designated as
healthcare
professionals i.e.
HCP) to perform their
routine fetal heart
ultrasound | The reference device
is a software that
aims at helping
sonographers,
OB/GYNs, MFMs and
Fetal surgeons (all
three designated as
healthcare
professionals i.e.
HCP) to perform their
routine fetal heart
ultrasound | FETOLY-HEART is a
software that aims at
helping
sonographers,
OB/GYNs, MFMs and
Fetal surgeons (all
three designated as
healthcare
professionals i.e.
HCPs) to perform
their routine fetal
heart ultrasound | Substantially
equivalent
The subject device
and the predicate
devices have the
same objective. |
| | examinations in real-time. | examinations in real-time. | examinations in real-time. | |
| Indications for
use | The predicate device
is intended to analyze
fetal ultrasound
images and clips
using machine
learning techniques
to automatically
detect heart views,
detect anatomical
structures within the
views and verify
quality criteria of the
views.
The device is
intended for use as a
concurrent reading
aid during the
acquisition and
interpretation of fetal
ultrasound images. | The device is a
general purpose
ultrasound system
intended for use by
qualified and trained
healthcare
professionals. | FETOLY-HEART is
intended to analyse
fetal ultrasound
images and clips
using machine
learning techniques
to automatically
detect heart views
and quality criteria
within the views. The
device is intended for
use as a concurrent
reading aid during
the acquisition and
interpretation of fetal
ultrasound images.
FETOLY-HEART is
indicated for use
during routine fetal
heart examination of
2nd and 3rd
trimester pregnancy
(gestational age:
from 17 to 40 weeks). | Substantially
equivalent
Indications for Use
are the same
between predicate
and subject devices. |
| Targeted
population | Pregnant women
during the 2nd and
3rd trimester of
pregnancy | Pregnant women
during the 2nd
trimester of
pregnancy | Pregnant women
during the 2nd and
3rd trimester of
pregnancy | Substantially
equivalent
Subject device has
the same intended
patient population
than the predicate
devices. |
| Clinical
outcome | - Images labeled with
correct fetal heart
view | - Images labeled
with correct fetal
heart view | - Images labeled with
correct fetal heart
view for patient cases | Substantially
equivalent
Performance testing
has successfully
validated the clinical
outcomes |
| | - Quality criteria
identified as
"Verified" when
detected and "Not
verified" when not
detected | - Quality criteria
identified as "Found"
when detected and
"Not found" when
not detected | - Quality criteria
identified as
"Verified" when
detected and "Not | Substantially
equivalent
In the subject device,
quality criteria
bounding box
localization can be |
| | | | verified" when not
detected
- Images labeled with
the localization of
quality criteria | optionally displayed
on post-acquisition
images to enhance
explainability of the
Al model.
Performance testing
has been performed
and does not
introduce new
questions of safety
and effectiveness. |
| Intended user | Qualified healthcare
professional
specialized in
prenatal ultrasound
imaging | Qualified healthcare
professional
specialized in
prenatal ultrasound
imaging | Qualified healthcare
professional
specialized in
prenatal ultrasound
imaging | Substantially
equivalent
Subject device has
the same intended
users as the predicate
devices. |
| Clinical
applications | Fetal/Obstetrics | Fetal/Obstetrics | Fetal/Obstetrics | Substantially
equivalent
Clinical application is
the same for subject
and predicate
devices. |
| Inclusion of a
PCCP | N/A | N/A | Included | Different
The PCCP in the
subject device
includes proposed
modifications related
to modifying model
training
hyperparameters,
additional retraining
with new training and
validation datasets
collected, and
addition/removal of
heart quality criteria. |
| Functionality 1: completeness overview | | | | |
| Automatically
detect views | Detection of 4ch, 3vx,
LVOT, RVOT and Abd
views (complete
implementation of
ISUOG
recommendations) | Detection of 4ch, 3vx,
LVOT, RVOT and Abd
views (complete
implementation of
ISUOG
recommendations) | Detection of 4ch, 3vx,
LVOT, RVOT and Abd
views (complete
implementation of
ISUOG
recommendations) | Substantially
equivalent
The subject device
includes the
detection of the same
views than the
predicate device |
| Automatically
detect quality
criteria | Detection of 28 heart
quality criteria. The
quality also
incorporates the
zoom level of the
view. | Detection of NA
heart quality criteria. | Detection of 52 heart
quality criteria (see
Table 1). The quality
also incorporates the
zoom level of the
view. | Substantially
equivalent
The subject device
includes the
detection of new
quality criteria when
compared to the
predicate device. This
quantitative
enhancement has
been tested and does
not raise any new
question of safety
and effectiveness. |
| Functionality 2: completeness illustration | | | | |
| Automatically
selects views | NA | Automatic suggestion
of views from a
sequence of images. | Automatic extraction
of views from a
sequence of images. | Substantially
equivalent
This image selection
functionality is a
feature absent in the
predicate device but
present in the
reference device.
Software testing has
been performed to
validate its use and
does not introduce
new questions of
safety and
effectiveness. |
| Technical characteristics | | | | |
| Data input | Accepts images and
image sequences
from ultrasound
machines | Accepts images and
image sequences
from ultrasound
machines | Accepts images and
image sequences
from ultrasound
machines | Substantially
equivalent
The input data is the
same for the subject
device and the
predicate and
reference devices. |
| Algorithm
Methodology | Artificial Intelligence:
Utilizes computer
vision algorithms to
analyze ultrasound
images and provides
visualization of
detected landmarks | Artificial Intelligence:
Utilizes computer
vision algorithms to
analyze ultrasound
images and provides
visualization of
detected landmarks | Artificial Intelligence:
Utilizes computer
vision algorithms to
analyze ultrasound
images and provides
visualization of
detected landmarks | Substantially
equivalent
The subject device
and the primary
predicate device use
both artificial
intelligence. |
| Platform | Operates in a cloud-
based environment
functioning
independently from
the ultrasound
equipment. | Operates as a local
software embedded
in the ultrasound
equipment. | Operates as a local
software functioning
independently from
the ultrasound
equipment. | Substantially
equivalent
The edge-based
approach reduces
exposure to potential
cloud-related
vulnerabilities and
latency issues.
Therefore, this
difference does not
raise any safety or
effectiveness
concerns. |
| Ultrasound
Machine
compatibility | Compatible with
ultrasound system
from GE Medical,
Samsung, Canon and
Philips | NA | Compatible with
ultrasound system
from GE Medical,
Samsung and Canon | Substantially
equivalent
This compatibility has
been tested and
validated as part of
device
generalizability in the
performance testing
study. |
| User
interaction | The user can interact
with the software to
override the
software's outputs.
The user has the
ability to review and
edit/override the
matching at any time
during or at the end
of the exam. | NA | The user can interact
with the software to
override the
software's outputs.
The user has the
ability to review and
edit/override the
matching at any time
during or at the end
of the exam. | Substantially
equivalent
User interactions are
the same between
primary predicate
and subject devices. |
9
Image /page/9/Picture/0 description: The image shows a logo for a company called DIAGNOLY. The logo features a blue magnifying glass with a stack of three curved lines inside, suggesting a focus on analysis or investigation. The company name, "DIAGNOLY," is written in a stylized blue font below the magnifying glass, emphasizing the company's diagnostic or analytical services.
10
Image /page/10/Picture/0 description: The image shows a blue magnifying glass with a symbol inside. The symbol consists of several horizontal lines that are stacked on top of each other, resembling a radar or sonar display. Below the magnifying glass, the word "DIAGNOLY" is written in blue, with the letters in a slightly rounded font.
11
Image /page/11/Picture/0 description: The image shows a logo for a company called DIAGNOLY. The logo features a blue magnifying glass with a stylized image of a stack of objects inside the lens. The word "DIAGNOLY" is written in blue letters below the magnifying glass.
12
Image /page/12/Picture/0 description: The image shows a blue logo for a company called DIAGNOLY. The logo features a magnifying glass with a stylized image of three stacked, truncated cones inside the lens. The word "DIAGNOLY" is written in blue below the magnifying glass.
Non-clinical performance data 7
7.1 FETOLY-HEART testing strategy
The following V&V testing were included into the development of the system:
- Software verification testing per IEC 62304 standard
- Tablet compatibility testing
- Cybersecurity verification testing ●
- Software Al model validation
13
Image /page/13/Picture/1 description: The image shows a logo for a company called "DIAGNOLY". The logo features a blue magnifying glass with a stylized image of a stack of three horizontal lines inside the lens. The word "DIAGNOLY" is written in blue below the magnifying glass.
FETOLY-HEART uses a machine learning (ML) algorithm for detection of heart views and quality criteria within these views in ultrasound images. Modifications to FETOLY-HEART will be made in accordance with the guiding principles on predetermined change control plans (PCCP) for machine learning-enabled medical devices. This PCCP provides a description of the device's planned modifications, and those modifications will be triggered and implemented in a controlled manner that ensures the continued safety and efficacy on the performance testing dataset, mitigating risks associated with changes to the ML model to not adversely impact the device's performance, safety, or effectiveness associated with its indications for use, and an impact assessment of the planned modifications.
In accordance with the PCCP, all algorithm modifications will be adequately trained, tuned, and locked prior to release of the software with the modified ML model. The PCCP does not include the implementation of adaptive algorithms that will continuously learn in the field. Implemented modifications to the FETOLY-HEART algorithm will be communicated to users via the software update notifications and through updated labelling. The modifications outlined in the PCCP are summarized in the table below. The PCCP in the subject device with the proposed modifications related to modifying model training hyperparameters, additional retraining and validation datasets collected, and addition/removal of heart quality criteria do not raise different questions of safety and effectiveness from the predicate device (see table below).
| Modification | Rationale | Testing Methods | Impact Assessment |
|---|---|---|---|
| Modification of | |||
| training and/or | |||
| validation datasets | Increase or recovery (in | ||
| case of data drift) of | |||
| FETOLY-HEART's | |||
| performance. | Re-training of the FETOLY- | ||
| HEART model with new | |||
| data to optimize its | |||
| performance followed by | |||
| internal testing and a | |||
| comparison of the initial | |||
| model to the modified | |||
| model using performance | |||
| metrics on the test | |||
| dataset. | Increased performance metrics of the | ||
| modified model for view or quality | |||
| criteria detection. | |||
| Benefits: Increase or recovery of | |||
| performance; generalization for | |||
| diverse cases. | |||
| Risks: Performance decrease | |||
| (overfitting, unintended bias). | |||
| Risk mitigation: The modified model | |||
| will be tested for superiority on the | |||
| performance study test dataset which | |||
| will contain new unseen data. | |||
| Modification of | |||
| model training | |||
| hyperparameters | Improvement and | ||
| optimization of FETOLY- | |||
| HEART's performance | Re-training of the FETOLY- | ||
| HEART model with new | |||
| parameters to optimize its | |||
| performance followed by | |||
| internal testing and a | |||
| comparison of the initial | |||
| model to the modified | |||
| model using performance | Increased performance metrics of the | ||
| modified model for view or quality | |||
| criteria detection. | |||
| Benefits: Increased performance; | |||
| generalization for diverse cases. | |||
| Risks: Performance decrease | |||
| (overfitting, unintended bias). |
Summary of changes to FETOLY-HEART per the PCCP:
14
Image /page/14/Picture/0 description: The image shows a blue logo for a company called "DIAGNOLY". The logo features a blue magnifying glass with a stack of three cone-shaped objects inside. The text "DIAGNOLY" is written in blue below the magnifying glass.
| | | metrics on the test
dataset. | Risk mitigation: The modified model
will be tested for superiority on the
performance study test dataset which
will contain new unseen data. |
|-----------------------------------------------|------------------------------------------------------------------------------------------------------------------------------------------------|---------------------------------------------------------------------------------------------------------------------------|------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| Heart quality
criteria
addition/removal | Maintain alignment
with quality criteria
recommended for fetal
heart screening in
state-of-the-art
international
guidelines. | New quality criteria will be
controlled using the same
acceptance criteria as
defined by secondary
endpoints. | Enhanced compliance with standard
international guidelines.
Benefits: Keeping the device relevant
by aligning with the updated list of
heart quality criteria.
Risks: Performance decrease and user
confusion.
Risk mitigation: Proper performance
testing with no decrease in test
performance. This change only
pertains to quality criteria belonging
to one of the 5 heart views already
included in FETOLY-HEART. |
FETOLY-HEART performance study 7.2
Diagnoly conducted a standalone performance testing on a dataset of 2,288 fetal ultrasound images across 480 patient cases, including full examination still images, cardiac clip frames and full examination video frames from 7 clinical sites in the United States and cases are representative of the intended use population. This testing dataset originated from distinct clinical sites from which the data used during model development (training/validation) was sourced, ensuring testing independence.
The results of the standalone performance testing demonstrate that FETOLY-HEART automatically detects fetal heart ultrasound views with a sensitivity ≥ 85% (acceptance criterion) and specificity ≥ 85% (acceptance criterion), detects quality criteria within heart views with a sensitivity ≥ 90% (acceptance criterion) and a specificity ≥ 90% (acceptance criterion), and localizes bounding boxes of quality criteria with a mean intersection over union (loU) of ≥ 50% (acceptance criterion). Sensitivity and specificity were evaluated individually for each view, and the performance goal to exceed 85% as the lower bound of the corresponding 95% Confidence Interval (Cl) was met. The Cl was estimated using bootstrap resampling at the subject level based on 1,000 samples, with traditional bootstrap Cl confidence limits derived as the 2.5th and 97.5th percentiles of the distribution of bootstrap estimates. The results are summarized in the tables below:
| Sensitivity | Specificity | |||||
|---|---|---|---|---|---|---|
| Fetal heart view | N(positive) | Point | ||||
| estimate | Bootstrap Cl | |||||
| (95%) | N(negative) | Point | ||||
| estimate | Bootstrap Cl | |||||
| (95%) | ||||||
| Abdomen view | 428 | 0.976 | (0.960,0.990) | 1860 | 0.998 | (0.996,1.000) |
15
Image /page/15/Picture/0 description: The image shows a logo for a company called "Diagnoly". The logo features a blue magnifying glass with a stylized image of a sound wave inside. The company name "Diagnoly" is written in blue, sans-serif font below the magnifying glass.
| Four chamber view | 391 | 0.987 | (0.974,0.997) | 1897 | 1.00 | (1.000,1.000) |
|---|---|---|---|---|---|---|
| Left ventricular outflow | ||||||
| tract view | 360 | 0.983 | (0.969,0.994) | 1928 | 0.999 | (0.998,1.000) |
| Right ventricular outflow | ||||||
| tract view | 313 | 0.987 | (0.974,0.996) | 1975 | 0.998 | (0.996,1.000) |
| Three vessels view | 316 | 0.981 | (0.965,0.993) | 1972 | 0.998 | (0.997,1.000) |
| Other view | 480 | 0.985 | (0.972,0.995) | 1808 | 0.983 | (0.977,0.989) |
| Sensitivity | Specificity | mloU | ||||||
|---|---|---|---|---|---|---|---|---|
| Quality criterion | N | Point estimate | Bootstrap Cl (95%) | N | Point estimate | Bootstrap CI (95%) | Point estimate | Bootstrap CI (95%) |
| Abdomen view | ||||||||
| Spine | 417 | 0.966 | (0.949,0.981) | 1871 | 0.995 | (0.992,0.998) | 0.739 | (0.728,0.750) |
| Left rib | 401 | 0.903 | (0.873,0.933) | 1887 | 0.997 | (0.995,0.999) | 0.651 | (0.631,0.673) |
| Right rib | 388 | 0.910 | (0.882,0.940) | 1900 | 0.997 | (0.994,0.999) | 0.624 | (0.602,0.643) |
| Descending aorta | 417 | 0.947 | (0.925,0.969) | 1871 | 0.998 | (0.996,0.999) | 0.528 | (0.515,0.542) |
| Inferior vena cava | 363 | 0.915 | (0.884,0.941) | 1925 | 0.992 | (0.988,0.996) | 0.512 | (0.496,0.528) |
| Stomach | 414 | 0.976 | (0.961,0.988) | 1874 | 0.995 | (0.992,0.998) | 0.734 | (0.720,0.746) |
| Umbilical vein | 345 | 0.962 | (0.942,0.980) | 1943 | 0.991 | (0.987,0.994) | 0.676 | (0.660,0.693) |
| Thorax apex | 397 | 0.919 | (0.893,0.944) | 1891 | 0.993 | (0.989,0.997) | 0.571 | (0.549,0.592) |
| Four chamber view | ||||||||
| Spine | 339 | 0.965 | (0.944,0.982) | 1949 | 0.999 | (0.997,1.000) | 0.768 | (0.754,0.779) |
| Left rib | 308 | 0.945 | (0.917,0.970) | 1980 | 0.996 | (0.993,0.999) | 0.744 | (0.726,0.764) |
| Right rib | 309 | 0.958 | (0.934,0.980) | 1979 | 0.996 | (0.993,0.998) | 0.749 | (0.730,0.768) |
| Descending aorta | 361 | 0.981 | (0.965,0.994) | 1927 | 0.996 | (0.994,0.999) | 0.646 | (0.632,0.659) |
| Left pulmonary vein | 127 | 0.921 | (0.871,0.965) | 2161 | 0.998 | (0.995,1.000) | 0.628 | (0.601,0.653) |
| Right pulmonary vein | 198 | 0.904 | (0.860,0.943) | 2090 | 0.996 | (0.993,0.999) | 0.601 | (0.578,0.623) |
| Left atrium | 387 | 0.990 | (0.979,0.997) | 1901 | 0.998 | (0.996,1.000) | 0.759 | (0.747,0.771) |
| Right atrium | 391 | 0.987 | (0.975,0.997) | 1897 | 1.00 | (1.000,1.000) | 0.774 | (0.763,0.786) |
| Foramen ovale flap | 113 | 0.929 | (0.883,0.971) | 2175 | 1.00 | (0.999,1.000) | 0.530 | (0.505,0.561) |
| Open Foramen Ovale | 348 | 0.951 | (0.928,0.972) | 1940 | 0.996 | (0.993,0.999) | 0.616 | (0.598,0.634) |
| Mitral valve | 207 | 0.908 | (0.866,0.948) | 2081 | 0.998 | (0.995,1.000) | 0.676 | (0.654,0.695) |
| Tricuspid valve | 243 | 0.959 | (0.931,0.983) | 2045 | 0.995 | (0.992,0.998) | 0.718 | (0.701,0.735) |
| Connection | ||||||||
| between | ||||||||
| crux and atrial septum | 299 | 0.943 | (0.915,0.969) | 1989 | 0.990 | (0.986,0.994) | 0.587 | (0.567,0.603) |
| Atrioventricular valve | ||||||||
| offset in crux | 115 | 0.939 | (0.897,0.980) | 2173 | 0.992 | (0.988,0.995) | 0.648 | (0.623,0.673) |
| Connection between | ||||||||
| interventricular | ||||||||
| septum and crux | 177 | 0.910 | (0.868,0.951) | 2111 | 0.996 | (0.993,0.999) | 0.565 | (0.544,0.585) |
16
Image /page/16/Picture/0 description: The image shows a blue logo for a company called Diagnoly. The logo features a magnifying glass with a stack of three objects inside. The text "DIAGNOLY" is written in blue below the magnifying glass.
| Interventricular
| septum | 216 0.917 | (0.877,0.949) | 2072 0.996 | (0.993,0.999) | 0.71 | (0.689,0.728) |
|---|---|---|---|---|---|---|
| Left ventricle | 389 0.987 | (0.975,0.997) | 1899 1.00 | (1.000,1.000) | 0.748 | (0.734,0.761) |
| Right ventricle | 387 0.990 | (0.979,0.997) | 1901 0.999 | (0.997,1.000) | 0.758 | (0.746,0.770) |
| Sternum | 322 0.972 | (0.952,0.988) | 1966 0.991 | (0.986,0.995) | 0.681 | (0.661,0.700) |
| Left ventricular outflow tract view | ||||||
| Left atrium | 356 0.98 | (0.966,0.992) | 1932 0.997 | (0.995,0.999) | 0.656 | (0.633,0.676) |
| Proximal ascending | ||||||
| aorta | 360 0.983 | (0.969,0.994) | 1928 0.999 | (0.998,1.000) | 0.716 | (0.702,0.730) |
| Semilunar valves | 165 0.909 | (0.866,0.950) | 2123 0.998 | (0.995,1.000) | 0.527 | (0.503,0.552) |
| Left ventricle | 358 0.978 | (0.961,0.992) | 1930 0.999 | (0.998,1.000) | 0.745 | (0.732,0.758) |
| Interventricular | ||||||
| septum | 199 0.925 | (0.888,0.959) | 2089 0.995 | (0.992,0.998) | 0.663 | (0.638,0.690) |
| Right ventricle | 358 0.980 | (0.964,0.992) | 1930 0.999 | (0.997,1.000) | 0.698 | (0.675,0.719) |
| Right ventricular outflow tract view | ||||||
| Descending aorta | 294 0.949 | (0.924,0.971) | 1994 0.997 | (0.994,0.999) | 0.617 | (0.595,0.638) |
| Trachea / bronchi | 180 0.922 | (0.882,0.960) | 2108 0.990 | (0.985,0.994) | 0.520 | (0.491,0.548) |
| Left pulmonary artery | 125 0.960 | (0.921,0.992) | 2163 0.995 | (0.992,0.998) | 0.615 | (0.582,0.646) |
| Ductus arteriosus | 177 0.944 | (0.907,0.974) | 2111 0.997 | (0.995,0.999) | 0.631 | (0.606,0.656) |
| Right | ||||||
| pulmonary | ||||||
| artery | 302 0.980 | (0.965,0.994) | 1986 0.994 | (0.991,0.997) | 0.654 | (0.633,0.675) |
| Origin of | ||||||
| pulmonary | ||||||
| arteries | 258 0.922 | (0.889,0.953) | 2030 0.997 | (0.994,0.999) | 0.589 | (0.568,0.610) |
| Septum | ||||||
| between | ||||||
| pulmonary | ||||||
| artery | ||||||
| trunk and | ||||||
| ascending | ||||||
| aorta | 218 0.945 | (0.911,0.973) | 2070 0.993 | (0.989,0.996) | 0.692 | (0.672,0.710) |
| Ascending aorta | 312 0.974 | (0.957,0.990) | 1976 0.998 | (0.996,1.000) | 0.633 | (0.613,0.650) |
| Superior vena cava | 218 0.940 | (0.910,0.971) | 2070 0.994 | (0.991,0.998) | 0.577 | (0.549,0.603) |
| Pulmonary trunk | 312 0.971 | (0.952,0.990) | 1976 0.999 | (0.998,1.000) | 0.760 | (0.744,0.775) |
| Three vessels views | ||||||
| Spine | 276 0.953 | (0.929,0.975) | 2012 0.998 | (0.996,1.000) | 0.792 | (0.779,0.805) |
| Trachea | 206 0.932 | (0.899,0.964) | 2082 0.997 | (0.994,0.999) | 0.539 | (0.517,0.562) |
| Side space on the left | ||||||
| of ductus | 270 0.933 | (0.901,0.961) | 2018 0.991 | (0.987,0.995) | 0.784 | (0.767,0.800) |
| Main | ||||||
| pulmonary | ||||||
| artery | 169 0.947 | (0.910,0.976) | 2119 0.999 | (0.998,1.000) | 0.733 | (0.708,0.755) |
| Ductus | 141 0.972 | (0.944,0.993) | 2147 0.998 | (0.996,1.000) | 0.749 | (0.724,0.771) |
| Ascending aorta | 181 0.978 | (0.954,0.995) | 2107 0.999 | (0.997,1.000) | 0.621 | (0.599,0.643) |
| Aortic arch | 133 0.97 | (0.942,0.993) | 2155 0.997 | (0.995,0.999) | 0.724 | (0.699,0.747) |
| Superior vena cava | 303 0.964 | (0.942,0.983) | 1985 0.996 | (0.994,0.999) | 0.528 | (0.508,0.547) |
| Thymus/sternum | 212 0.915 | (0.876,0.953) | 2076 0.998 | (0.995,1.000) | 0.716 | (0.693,0.737) |
17
Image /page/17/Picture/1 description: The image shows a logo for a company called DIAGNOLY. The logo features a blue magnifying glass with a stack of three cone-shaped objects inside the lens. The word "DIAGNOLY" is written in blue below the magnifying glass.
The performance validation dataset included the following clinical subgroups: maternal age, gestational age, territory, BMI, scanner manufacturer, heart abnormality. It also comprised races and ethnicities and various clinical sites, ensuring representation across the intended use population. Image digital quality (bad, average, good) and image type (still image from entire examination, cardiac clip frame, full examination frame) were identified as potential controlled for. Performance metrics were analyzed for each subgroup and confounder to validate the model's robustness and generalizability. The subgroups distribution is summarized in the table below:
| Number of cases | Number of images | ||
|---|---|---|---|
| Subgroup | |||
| (total=480) | (total=2,288) | ||
| Center 1 | 99 (20.6%) | 505 (22.1%) | |
| Center | Center 2 | 63 (13.1%) | 248 (10.8%) |
| Center 3 | 84 (17.5%) | 378 (16.5%) | |
| Center 4 | 55 (11.5%) | 264 (11.5%) | |
| Center 5 | 54 (11.2%) | 275 (12.0%) | |
| Center 6 | 55 (11.5%) | 265 (11.6%) | |
| Center 7 | 70 (14.6%) | 353 (15.4%) | |
| Territory | US | 234 (48.8%) | 1157 (50.6%) |
| EU | 246 (51.2%) | 1131 (49.4%) | |
| Gestational age | 2nd trimester | 233 (48.5%) | 1189 (52.0%) |
| 3rd trimester | 243 (50.6%) | 1075 (47.0%) | |
| Unknown | 4 (0.8%) | 24 (1.0%) | |
| Maternal age |