Mayo Clinic is a nonprofit academic medical center headquartered in Rochester, Minnesota, with additional campuses in Arizona and Florida and a regional health system spanning three Upper Midwest states. The organization employed nearly 85,000 people in 2025 and posted $473 million in operating income. Mayo committed $9 billion in capital investment through its “Bold. Forward. Unbound.” expansion. Independent rankings have named it Newsweek’s No. 1 hospital in the world for the eighth consecutive year. This scale extends to artificial intelligence, with Mayo describing over 200 AI projects in various stages. In 2025, Mayo integrated 22 AI-enabled Mayo Clinic Platform solutions into clinical practice and secured nearly 200 new AI, biopharma, and diagnostics agreements. Since its 2019 launch, the Mayo Clinic Platform has built a research data infrastructure with over 15 million patient records and billions of radiology images, lab results, and clinical notes. This article explores two AI use cases illustrating how Mayo applies its data and clinical expertise: AI-Enabled ECG Screening for Early Disease Detection and AI-Powered Chart Review with Record Time.
AI-Enabled ECG Screening
Asymptomatic left ventricular dysfunction, a precursor to heart failure, often goes undetected by standard screening, traditionally requiring an echocardiogram. With heart failure affecting millions of Americans and incurring significant costs, Mayo's cardiology researchers sought an early detection method using existing data. They screened over 625,000 paired ECG and echocardiogram records, training a neural network on nearly 98,000 pairs to identify electrical patterns in standard EKGs associated with a weakened heart pump. This AI approach has expanded to flag other conditions like atrial fibrillation, cardiac amyloidosis, aortic stenosis, hypertrophic cardiomyopathy, and biological age, working across both traditional 12-lead ECGs and single-lead smartwatch/digital stethoscope readings. The amyloidosis version was validated across 25,525 patients in four US health systems, showing 78.9% sensitivity and 91.2% specificity. The model's quality relies on Mayo's extensive historical data. This screening adds no new burden to clinicians or patients, as the AI analyzes already collected ECGs. A positive result prompts a confirmatory echocardiogram or referral. Portable versions further extend screening beyond the clinic, as demonstrated by an AI-enabled digital stethoscope flagging twice as many cases of peripartum cardiomyopathy in a Nigerian study compared to standard care, highlighting AI's adoption success when integrated into existing workflows. The tool was prospectively tested in the EAGLE trial, involving 22,641 patients across 348 primary care clinicians and 45 medical centers. Results, published in Nature Medicine, showed AI-guided screening increased low ejection fraction diagnoses by 32% (about five additional diagnoses per 1,000 patients). The 12-lead algorithm is FDA-cleared and licensed to Anumana (a company co-founded by Mayo with nference), while Mayo partnered with Eko Health for a single-lead version. Newer applications, like amyloidosis detection (FDA-cleared in April 2026), are in early commercial rollout. This is a robust, evidence-backed AI application, not an experimental pilot.
AI-Powered Chart Review with Record Time
Mayo Clinic physicians frequently face the daunting task of reviewing hundreds of pages of fragmented medical records from other health systems before patient visits, especially for those seeking second or third opinions. Dr. Alexander Ryu, an internal medicine physician and vice chair of innovation, noted that Mayo receives tens of millions of pages annually, necessitating a solution to extract crucial information. This issue is widespread; a 2025 survey in the Journal of the American Medical Informatics Association indicated that immature AI tools are a major barrier to adoption, underscoring the value of administrative AI solutions over broad diagnostic claims. To address this, Mayo developed Record Time with Scale AI, leveraging the Scale Generative AI Platform. This tool ingests and chronologically organizes external patient records, producing concise summaries and making the information searchable. Beyond chart review, the collaboration also aims to automate the detection of safety events, such as wrong-site surgeries or falls, often obscured within routine reporting. All data utilized in this initiative remains within Mayo Clinic's HIPAA-compliant environment. This AI tool allows physicians to reduce manual patient history assembly, thereby allocating more time to direct patient interaction. Scale AI reports an average gain of 11 more minutes per patient, while Dr. Ryu observed time savings of five to thirty minutes per visit, depending on case complexity, and confirmed the tool's efficacy in preventing oversight of critical details. Record Time is not a pilot but a deployed, in-use tool, one of approximately 150 AI models running across Mayo’s system, demonstrating its proven utility beyond a single department or trial.
