Bottom line
Anumana, a Mayo Clinic and nference joint venture based in Cambridge, Massachusetts, sells ECG-AI algorithms that read a standard 12-lead ECG and send a clinical decision support notification into the EHR. It holds four 510(k) clearances for low ejection fraction, pulmonary hypertension and cardiac amyloidosis. It suits a health system that wants those flags raised from 12-lead ECGs it already acquires. Access is by demo request, so a practice that wants a posted price or a device to buy off the page is not served.
ECG-AI software that flags low ejection fraction, pulmonary hypertension and cardiac amyloidosis from standard 12-lead ECGs.
No public price. The pricing page returns 404; the site offers a demo request only.
Which FDA decisions are Anumana's?
All four are 510(k) clearances:
- K232699 (September 28, 2023), Low Ejection Fraction AI-ECG Algorithm, product code QYE. Its predicate is Viz HCM, authorized through De Novo DEN230003 for Viz.ai, which is another company's authorization.
- K250652 (July 28, 2025), ECG-AI Low Ejection Fraction 12-Lead algorithm.
- K252360 (March 28, 2026), ECG-AI Pulmonary Hypertension 12-Lead algorithm, product code SAT.
- K253801 (April 7, 2026), ECG-AI Cardiac Amyloidosis 12-Lead algorithm, product code SHP. The vendor's site dates it April 8.
Where do the algorithms appear in the literature?
Thirteen PubMed hits match the company by name or affiliation. Nine concern its work, many with Mayo and nference authors, which makes them vendor-linked. Titles name these, and whether each one tests the cleared algorithm is not verified:
- PMID 41377035 (Am Heart J Plus, 2026), a pragmatic trial of the Anumana clinical decision support tool for guideline-directed management, with PMID 42828092 as a comment on it.
- PMID 40276542 (Am Heart J Plus, 2025), a multicenter pragmatic implementation of AI-ECG software to identify low LVEF.
- PMID 41547169 (JACC Advances, 2026), multisite external validation of an AI-ECG algorithm for low EF.
- PMID 38936966 (European Respiratory Journal, 2024), ECG-based AI for early pulmonary hypertension detection.
The vendor says the low EF model was built on about 2.9 million ECG and echo pairs from more than 676,000 patients and reports 90.2% sensitivity and 85.1% specificity, attributed to its own data. It claims more than 30 peer-reviewed publications.
How does it fit into a hospital?
The vendor says the software accepts XML or DICOM waveform files, that no PHI is exported for external processing, and that the algorithms deploy within EHRs. No EHR or ECG management system is named. The homepage lists CPT III codes for charge capture and calls the product "eligible for reimbursement today". It describes low EF as available for clinical use in the US and EU and cites no CE certificate. Beyond a vulnerability disclosure page, the homepage and About page cite no security certification, and they list no price or customer.
Ownership and funding
The company launched in 2021 as a joint venture of Mayo Clinic and nference. Maulik Nanavaty is CEO and Murali Aravamudan executive chairman. MedCity News coverage reports a $25.7 million Series A at launch and a $60 million Series B by July 2022. CB Insights shows a Series C and $153.1 million in total, which is unverified.
The verdict
The homepage cites CPT III codes for charge capture and calls the product eligible for reimbursement. A buyer who wants single-lead or wearable ECG screening is in a different category.
Editorial review last generated November 2, 2026. Synthesized from clinician sentiment, peer-reviewed coverage, and our editorial silo picks. Refined by hand where vendor facts change.





