Low Ejection Fraction AI-ECG Algorithm

K232699 · Anumana, Inc. · QYE · Sep 28, 2023 · Cardiovascular

Device Facts

• Record ID: K232699
• Device Name: Low Ejection Fraction AI-ECG Algorithm
• Applicant: Anumana, Inc.
• Product Code: QYE · Cardiovascular
• Decision Date: Sep 28, 2023
• Decision: SESE
• Submission Type: Traditional
• Regulation: 21 CFR 870.2380
• Device Class: Class 2
• Attributes: AI/ML, Software as a Medical Device, PCCP

Intended Use

The Anumana Low Ejection Fraction AI-ECG Algorithm is software intended to aid in screening for Left Ventricular Ejection Fraction (LVEF) less than or equal to 40% in adults at risk for heart failure. This population includes, but is not limited to: patients with cardiomyopathies patients who are post-myocardial infarction patients with aortic stenosis patients with chronic atrial fibrillation patients receiving pharmaceutical therapies that are cardiotoxic, and postpartum women. Anumana Low Ejection Fraction AI-ECG Algorithm is not intended to be a stand-alone diagnostic device for cardiac conditions, should not be used for monitoring of patients, and should not be used on ECGs with a paced rhythm. A positive result may suggest the need for further clinical evaluation in order to establish a diagnosis of Left Ventricular Ejection Fraction (LVEF) less than or equal to 40%. Additionally, if the patient is at high risk for the cardiac condition, a negative result should not rule out further non-invasive evaluation. The Anumana Low Ejection Fraction AI-ECG Algorithm should be applied jointly with clinician judgment.

Device Story

Software-as-a-Medical-Device (SaMD) analyzing 10-second 12-lead ECG digital waveforms; utilizes AI algorithm to predict likelihood of LVEF ≤ 40%. Operates as spot-check tool in primary, urgent, and emergency care settings; integrated into EMR or ECG management systems. Outputs 'Low LVEF Detected', 'Low LVEF Not Detected', or 'Error' to third-party software (smartphone, tablet, PC). Assists clinicians in identifying patients requiring further cardiac evaluation (e.g., echocardiography). Benefits patients by enabling early detection of low LVEF in settings where imaging is unavailable or difficult to access.

Clinical Evidence

Retrospective study of 16,000 patient-ECG pairs across 4 US health systems. Inclusion: age ≥ 18, digital 12-lead ECG paired with TTE (LVEF) within 30 days. Results: Sensitivity 84.5% (95% CI 82.2-86.6%), Specificity 83.6% (95% CI 82.9-84.2%), PPV 30.5%, NPV 98.4%. Subgroup analysis showed performance heterogeneity by age and BMI, but no significant differences by site, sex, race, or ECG device.

Technological Characteristics

SaMD packaged in Docker container; integrates with EMR/EMS. Inputs: 12-lead ECG digital waveform (500Hz). Outputs: Binary classification (Low LVEF detected/not detected). ML-based algorithm with Predetermined Change Control Plan (PCCP). No GUI; relies on third-party display systems.

Indications for Use

Indicated for adults at risk for heart failure, including those with cardiomyopathies, post-myocardial infarction, aortic stenosis, chronic atrial fibrillation, cardiotoxic therapy, or postpartum. Contraindicated for patients with paced rhythms.

Regulatory Classification

Identification

Viz HCM is a cardiovascular machine learning-based notification software intended to be used in parallel to the standard of care to analyze 12-lead ECG recordings from patients 18 years of age or older. It detects signs associated with hypertrophic cardiomyopathy (HCM) and allows the user to view the ECG and analysis results. It is not intended for use on patients with implanted pacemakers, does not replace standard diagnostic methods, and is not intended to rule out HCM or be used in lieu of a full patient evaluation.

Special Controls

In combination with the general controls of the FD&C Act, cardiovascular machine learningbased notification software is subject to the following special controls:

Predicate Devices

• Viz HCM (DEN230003)

Reference Devices

• Hemosphere Advanced Monitoring Platform and ClearSight Module (K201446)
• WorkMate Claris System (K210392)

Submission Summary (Full Text)

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

Image /page/0/Picture/0 description: The image shows the logo for the U.S. Food & Drug Administration (FDA). The logo consists of two parts: the Department of Health & Human Services seal on the left and the FDA acronym followed by the full name of the agency on the right. The FDA part of the logo is in blue, with the acronym in a solid blue square and the agency name in a lighter blue. The text reads "FDA U.S. FOOD & DRUG ADMINISTRATION".

September 28, 2023

Anumana, Inc. % Alexia Haralambous Senior Principal ROM+ 2251 San Diego Ave, Suite B-257 San Diego, California 92110

Re: K232699

Trade/Device Name: Low Ejection Fraction AI-ECG Algorithm Regulation Number: 21 CFR 870.2380 Regulation Name: Cardiovascular Machine Learning-Based Notification Software Regulatory Class: Class II Product Code: OYE Dated: September 3, 2023 Received: September 5, 2023

Dear Alexia Haralambous:

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

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

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

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

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

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

For comprehensive regulatory information about 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 ([e-mail]) or phone [phone] or 301-796-7100).

Sincerely,

# Robert T. Kazmierski -S

for

LCDR Stephen Browning Assistant Director Division of Cardiac Electrophysiology, Diagnostics, and Monitoring Devices Office of Cardiovascular Devices Office of Product Evaluation and Quality Center for Devices and Radiological Health

Enclosure

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

### Indications for Use

510(k) Number (if known) K232699

Device Name

Low Ejection Fraction AI ECG Algorithm

Indications for Use (Describe)

The Anumana Low Ejection Fraction AI-ECG Algorithm is software intended to aid in screening for Left Ventricular Ejection Fraction (LVEF) less than or equal to 40% in adults at risk for heart failure. This population includes, but is not limited to:

· patients with cardiomyopathies

- patients who are post-myocardial infarction
- · patients with aortic stenosis
- · patients with chronic atrial fibrillation
- · patients receiving pharmaceutical therapies that are cardiotoxic, and

• postpartum women.

Anumana Low Ejection Fraction Al-ECG Algorthm is not intended to be a stand-alone diagnostic device for cardiac conditions, should not be used for monitoring of patients, and should not be used on ECGs with a paced rhythm.

A positive result may suggest the need for further clinical evaluation in order to establish a diagnosis of Left Ventricular Ejection Fraction (LVEF) less than or equal to 40%. Additionally, if the patient is at high risk for the cardiac condition, a negative result should not rule out further non-invasive evaluation.

The Anumana Low Ejection Fraction AI-ECG Algorithm should be applied jointly with clinician judgment.

| Type of Use (Select one or both, as applicable) | |
|----------------------------------------------------------------------------------|---------------------------------------------------------------------------------|
| <div> <span> Prescription Use (Part 21 CFR 801 Subpart D) </span> </div> | <div> <span> Over-The-Counter Use (21 CFR 801 Subpart C) </span> </div> |

#### 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}------------------------------------------------

Image /page/3/Picture/0 description: The image shows the logo for Anumana. The logo consists of a stylized graphic to the left of the word "anumana". The graphic is made up of a dot and two vertical lines of varying heights. The word "anumana" is written in a sans-serif font.

# 510(k) Summary Low Ejection Fraction AI ECG Algorithm K232699

| Applicant Name: | Anumana, Inc.<br>One Main Street, Suite 400<br>East Arcade 4th Floor<br>Cambridge, MA 02142 |
|----------------------------------------------------------------------------|----------------------------------------------------------------------------------------------------------------------------------------------|
| Contact Person: | Suzanne Goodman<br>VP Regulatory and Quality<br>Anumana, Inc.<br>Tel. [phone]<br>[e-mail] |
| Consultant: | Alexia Haralambous, MS, RAC<br>Allison Komiyama, Ph.D, RAC<br>Erin Gontang, Ph.D.<br>RQM+<br>[e-mail]<br>Tel. [phone] |
| Date Summary Prepared: | September 22, 2023 |
| Proprietary Name of Device: | Low Ejection Fraction AI ECG Algorithm |
| Common or Usual Name: | ECG AI analysis tool |
| Classification Panel: | Cardiology |
| Regulation Number: | 21 CFR § 870.2380 |
| Regulation Name: | Cardiovascular machine learning-based notification software |
| Regulation Class: | Class II |
| Product Code: | QYE |
| 1 Predicate Device<br>Device Name:<br>Manufacturer:<br>Application Number: | VIZ HCM<br>Viz.ai<br>DEN230003 |

#### ನ Device Description

The Low Ejection Fraction AI-ECG Algorithm interprets 12-lead ECG voltage times series data using an artificial intelligence-based algorithm. The device analyzes 10 seconds of a single 12lead ECG acquisition, and within seconds provides a prediction of likelihood of LVEF (ejection fraction less than or equal to 40%) to third party software. The results are displayed by the thirdparty software on a device such as a smartphone, tablet, or PC. The Low Ejection Fraction AI-ECG Algorithm was trained to predict Low LVEF using positive and control cohorts, and the prediction of Low LVEF in patients is generated using defined conditions and covariates. The Low Ejection Fraction AI-ECG Algorithm device is intended to address the unmet need for a point-of-care screen for LVEF less than or equal to 40% and is expected to be used by

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

cardiologists, front-line clinicians at primary care, urgent care, and emergency care settings, where cardiac imaging may not be available or may be difficult or unreliable for clinicians to operate. Clinicians will use the Low Eiection Fraction AI-ECG Algorithm to aid in screening for LVEF less than or equal to 40% and making a decision for further cardiac evaluation.

#### 3 Intended Use/Indications for Use

The Anumana Low Ejection Fraction AI-ECG Algorithm is software intended to aid in screening for Left Ventricular Ejection Fraction (LVEF) less than or equal to 40% in adults at risk for heart failure. This population includes, but is not limited to:

- patients with cardiomyopathies ●
- . patients who are post-myocardial infarction
- . patients with aortic stenosis
- . patients with chronic atrial fibrillation
- patients receiving pharmaceutical therapies that are cardiotoxic, and .
- postpartum women.

Anumana Low Ejection Fraction AI-ECG Algorithm is not intended to be a stand-alone diagnostic device for cardiac conditions, should not be used for monitoring of patients, and should not be used on ECGs with a paced rhythm.

A positive result may suggest the need for further clinical evaluation in order to establish a diagnosis of Left Ventricular Ejection Fraction (LVEF) less than or equal to 40%. Additionally, if the patient is at high risk for the cardiac condition, a negative result should not rule out further non-invasive evaluation.

The Anumana Low Ejection Fraction AI-ECG Algorithm should be applied jointly with clinician judgment.

#### ব Technological Characteristics

The subject device, Anumana's Low Ejection Fraction AI-ECG Algorithm, is a Software as a Medical Device (SaMD) provided as a software module packaged in a Docker container. The Low Ejection Fraction AI-ECG Algorithm does not provide a graphical user interface (GUI) of its own. It is integrated with other medical systems such as Electronic Medical Record (EMR) systems or ECG Management Systems (EMS). The third-party integrating software furnishes a 12-lead ECG digital waveform as input to the Low Ejection Fraction AI-ECG Algorithm and records the algorithm output for display via the integrated medical system or for printing in an offline report.

#### Summary of Clinical Performance Data 5

The performance characteristics for the Low Ejection AI-ECG Algorithm have been clinically validated for the presence of a low ejection of less than or equal to 40% in patients, with a sensitivity and specificity greater than the study's predetermined acceptance criteria (80%).

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

| Performance Characteristic | Value |
|----------------------------|---------------|
| Sensitivity | 80% or higher |
| Specificity | 80% or higher |

### Table 1: Performance Characteristics

#### Algorithm Training

To develop the Anumana Low Ejection Fraction AI-ECG (Low EF ECG-AI) algorithm, 93,722 patients with an ECG and transthoracic echocardiogram (TTE) performed within a 2-week interval were identified from a research use authorized clinical database from Mayo Clinic. For each patient, the left ventricular ejection fraction (LVEF) measurement from the earliest TTE was selected and paired with the closest ECG recording within a 2-week interval of the TTE. The disease cohort was defined as LVEF ≤ 40% and consisted of 11% of the overall cohort. The control cohort was defined as LVEF > 40% and consisted of 89% of the overall cohort. The cohorts were split into training (50%), tuning (20%) and set-aside testing (30%) datasets. Within the set-aside test set, the sensitivity and specificity of the algorithm were 85.7% and 84.2%, respectively.

Performance analyses were also conducted in subgroups of age, sex, and race. Within the training set-aside test dataset, 91.2% of patients were White, 1.4% Black/African American, 1.3% Asian, 0.4% American Indian and 5.7% were of other or unknown races. Patients were 57.6% male and 43.4% female. The average age was 61.5 (± 16.4) years and 11.2% of patients were under age 40, 10.0% were 40-49, 18.1% were 50-59, 25.9% were 70-79, and 12.5% were over 80.

### Summary of Clinical Validation

### Study Design

The performance profile of the Anumana Low Ejection Fraction AI-ECG (Low EF AI-ECG) algorithm was validated in a retrospective study of 16,000 patient records across 4 health systems across the United States and 16,000 patients. The objective of the study was to establish the diagnostic performance of the device for the purpose of detecting the presence of an EF ≤ 40% in a clinically and demographically diverse population. The sole inclusion criteria for a subject was age ≥ 18 and the availability of at least one digital 12-lead ECG paired with an echocardiogram with quantitative left ventricular ejection fraction (LVEF) information within 30 days following the date of the ECG. Candidate cases were identified in reverse chronological order from the date the site was activated. For each patient, the most recent echocardiogram was paired with the most recent ECG (without pacing present) acquired for that patient prior to the echocardiogram, up to the 30 days limit. Four models of 12-L ECGs (GE Dash 3000, GE MAC® 5500 Resting Analysis System, GE MAC VU360 Resting ECG, and Philips PageWriter TC70 ECG machines (model 860315) were represented in the ECG inputs to the algorithm.

### Summary of Study Results

Each of the 4 sites contributed 4,000 patient-ECG pairs to a final pool of 16,000 patient-ECG pairs. The study sample was representative of the US population and was 65.9% White, 11.0%

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

Hispanic, 10.1% Black/African American, 5.3% American Indian or Alaska Native, 2.2% Asian, and 0.4% Native Hawaiian or Pacific Islander. The sample consisted of 52% male and 48% female participants. The average age as 66 and 11.6% of participants were under age 40, 9.2% were 40-49, 15.8% were 50-59, 22.8% were 60-69, 21.6% were 70-79, and 19% were over 80. A total of 2,040 records were excluded as a result of the quality checks performed by the Anumana Low Ejection Fraction AI-ECG (Low EF AI-ECG) algorithm.

Within this diverse study sample, a total of 1,096 LVEF ≤ 40% cases were identified (prevalence of 7.9%) from 13,960 samples. The Anumana Low EF ECG-AI device achieved a sensitivity of 84.5% (95% CI of 82.2% to 86.6%), a specificity of 83.6% (95% CI of 82.9% to 84.2%), a positive predictive value of 30.5% (PPV, 95% CI of 28.8% to 32.1%), and a negative predictive value of 98.4% (NPV, 95% CI of 98.2% to 98.7%). Results are provided in Figure 1: Matrix of primary results with subgroup analyses results provided in Table 2: Subgroup Analyses.

Image /page/6/Figure/4 description: This image shows a confusion matrix with the predicted values on the y-axis and the truth values on the x-axis. The top left quadrant shows 2,122 false positives, and the top right quadrant shows 928 true positives. The bottom left quadrant shows 10,848 true negatives, and the bottom right quadrant shows 170 false negatives. The specificity is 83.6%, the sensitivity is 84.5%, the PPV is 30.4%, and the NPV is 98.5%.

Figure 1: Matrix of primary results

#### Subgroup Analyses

Subgroup assessments of diagnostic performance were conducted to determine if there was heterogeneity in device performance across clinical sites, demographics, clinical characteristics, conduction disorders, ECG manufacturer, and ECG devices. To assess for heterogeneity, Breslow-Day tests for the strata were conducted on the odds ratio for condition presence given a positive ECG-AI result. The results are summarized in Table 2 below.

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

| Subgroup<br>Analysis | Result of Test for Heterogeneity |
|-----------------------------|------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| Clinical Site | Not statistically significant |
| Biological Sex | Not statistically significant |
| Race/Ethnicity | Not statistically significant |
| Age Group | Diagnostic performance varied across age strata ( $p < 0.01$ ). The diagnostic odds<br>ratio was higher than the overall estimate in patients aged 40 to 69 years and lower<br>in younger and older patients. In younger patients, sensitivity was lower while<br>specificity was higher, whereas in patients of advanced age, sensitivity was higher<br>while specificity was lower. |
| Clinical<br>Characteristics | Diagnostic performance varied across body mass index strata ( $p = 0.02$ ), with the lowest diagnostic performance observed in underweight (BMI $<18.5$ kg/m2) participants and the highest diagnostic performance observed in overweight (BMI 25+ to <30) and obese (BMI 30+) participants. Diagnostic performance varied across certain elements of medical history derived from ICD9/ICD10 codes in patient medical records. The diagnostic odds ratio was lower in patients with prior history of heart failure ( $p<0.01$ ), myocardial infarction ( $p<0.01$ ) and coronary revascularization ( $p=0.07$ ). In each of these cases, the disease prevalence was high (>17% for the three conditions) and the algorithm had higher sensitivity and lower specificity. This combination, at the higher prevalence values, resulted in robust positive predictive value (>35%) and negative predictive value (>94%). |
| Conduction<br>Disorders | Not statistically significant |
| ECG Manufacturer | Not statistically significant |
| ECG Device | Not statistically significant |

### Table 2: Subgroup Analyses

#### Summary of Non-Clinical Performance Data 6

The performance characteristics for the Low Ejection AI-ECG Algorithm have been evaluated with the following non-clinical testing: software verification and validation (per IEC 62304), cybersecurity, labeling validation, and human factors.

# Substantial Equivalence Conclusion

The subject device, Anumana Low Ejection Fraction AI-ECG Algorithm, is substantially equivalent to the predicate device Viz HCM (DEN230003). The devices have similar intended uses, principles of operation, and technical characteristics. Where differences occur between the subject device and the predicate, results of clinical performance and results of non-clinical verification and validation demonstrate that the subject device is as safe and as effective as the predicate.

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

| | Subject Device | Predicate Device | |
|-----------------------------------------|-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| Product Name | Anumana Low Ejection Fraction<br>AI-ECG Algorithm | Viz HCM | Comparison |
| Application No. | K232699 | DEN230003 | - |
| Product Codes | QXX<br>DQK | QXO | - |
| Regulation No. | 21 CFR 870.2380 | 21 CFR 870.2380<br>21 CFR 870.1425 | - |
| Regulation Name | Cardiovascular machine learning-<br>based notification software | Cardiovascular machine learning-<br>based notification software | - |
| Rx / OTC | Rx | Rx | Same |
| Intended Use/<br>Indications for<br>Use | The Anumana Low Ejection Fraction<br>AI-ECG Algorithm is software<br>intended to aid in screening for Left<br>Ventricular Ejection Fraction<br>(LVEF) less than or equal to 40% in<br>adults at risk for heart failure. This<br>population includes, but is not<br>limited to:<br>• patients with cardiomyopathies<br>• patients who are post-myocardial<br>infarction<br>• patients with aortic stenosis<br>• patients with chronic atrial<br>fibrillation<br>• patients receiving pharmaceutical<br>therapies that are cardiotoxic, and<br>• postpartum women.<br>Anumana Low Ejection Fraction AI-<br>ECG Algorithm is not intended to be<br>a stand-alone diagnostic device for | Viz HCM, a standalone<br>electrocardiogram (ECG) analysis<br>software to identify patients 18 or<br>older for further follow-up for<br>hypertrophic cardiomyopathy<br>(HCM), a rare condition in which the<br>heart muscle becomes abnormally<br>thick and makes it difficult for the<br>heart to pump blood. This device can<br>analyze recordings from compatible<br>12-lead ECGs, detect signs<br>associated with HCM, and allow the<br>user to view the ECG and software<br>analysis. The device does not provide<br>a diagnosis of HCM and is not<br>intended for use on patients with<br>implanted pacemakers | Similar – The subject device is<br>intended to aid in screening for Left<br>Ventricular Ejection Fraction<br>(LVEF) less than or equal to 40% in<br>adults at risk for heart failure,<br>whereas the predicate device is an<br>analysis software used to identify<br>patients for further follow-up for<br>hypertrophic cardiomyopathy<br>(HCM). |
| | Subject Device | Predicate Device | |
| Product Name | Anumana Low Ejection Fraction<br>AI-ECG Algorithm | Viz HCM | Comparison |
| | cardiac conditions, should not be<br>used for monitoring of patients, and<br>should not be used on ECGs with a<br>paced rhythm.<br>A positive result may suggest the<br>need for further clinical evaluation in<br>order to establish a diagnosis of Left<br>Ventricular Ejection Fraction<br>(LVEF) less than or equal to 40%.<br>Additionally, if the patient is at high<br>risk for the cardiac condition, a<br>negative result should not rule out<br>further non-invasive evaluation.<br>The Anumana Low Ejection Fraction<br>AI-ECG Algorithm should be<br>applied jointly with clinician<br>judgment. | | |
| Operational<br>Mode | Spot check | Spot check | Same |
| Motion | None | Hemosphere Advanced Monitoring<br>Platform and ClearSight Module<br>(K201446)<br>WorkMate Claris System (K210392) | Similar – The subject device is a<br>SaMD that does not need to take into<br>account motion correction. |
| Patient<br>Population | Adults | Adults | Same |
| Environment of<br>Use | Primary care, urgent care, and<br>emergency care settings | Hospital and Pre-Hospital Setting | Similar – Both devices are to be used<br>in professional healthcare settings |
| | Subject Device | Predicate Device | |
| Product Name | Anumana Low Ejection Fraction<br>AI-ECG Algorithm | Viz HCM | Comparison |
| Algorithm | Machine learning based algorithm<br>with a predetermined change control<br>plan (PCCP) | Machine learning based algorithm | Similar - The subject device<br>submission proactively pre-specifies<br>intended modifications (and their<br>method of implementation) that do<br>not require additional marketing<br>submission. |
| Algorithm<br>Calculation and<br>Output | Detection of LVEF (Left Ventricular<br>Ejection Fraction less than or equal<br>to 40%) from an ECG signal | Presence of signs associated with<br>HCM derived from an ECG signal | Similar - Both devices identify a<br>single cardiovascular condition |
| Ground Truth for<br>Model Training | Transthoracic echocardiogram (TTE)<br>with disease | Cardiologist adjudicated signs<br>associated with a disease using an<br>ECG waveform | Similar – Both devices rely upon<br>established clinical diagnostic<br>methods as ground truth |
| Physiological<br>Parameter Inputs | 12-Lead ECG waveform in digital<br>format | Parameters gathered from the<br>Hemosphere Advanced Monitoring<br>Platform and ClearSight Module<br>(K201446) and<br>WorkMate Claris System (K210392).<br>Includes oximetry, heart rate, and<br>additional cardiac parameters (e.g.,<br>blood pressure) using an EKG. | Similar - Both devices utilize ECG<br>digital waveforms as input |
| Data Displayed | Algorithm output is provided to third<br>party software that displays the result<br>to clinicians. Output provided for<br>each ECG is "Low LVEF Detected"<br>"Low LVEF Not Detected" or<br>"Error"). | Device flags suspected ECGs and<br>categorizes patients as requiring<br>further follow up. | Similar – Both devices provide data<br>suggesting the likelihood of a<br>cardiovascular disease or condition<br>for further referral or diagnostic<br>follow-up. |
| Hardware | Compatible 12-Lead diagnostic ECG<br>machines with 500Hz digital output | Hemosphere Advanced Monitoring<br>Platform and ClearSight Module<br>(K201446) and<br>WorkMate Claris System (K210392) | Similar - Both devices are<br>compatible with devices providing<br>digital ECG outputs |
| | Subject Device | Predicate Device | |
| Product Name | Anumana Low Ejection Fraction | Viz HCM | Comparison |
| | AI-ECG Algorithm | | |
| Software | Anumana proprietary algorithm and<br>application | Viz.AI proprietary algorithm and<br>application | Similar – Both devices incorporate<br>their own proprietary algorithms and<br>applications. |

### Table 3: Substantial Equivalence Comparison of Subject Device to the Predicate

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

{10}------------------------------------------------

{11}------------------------------------------------