The AI tool represents a quantum leap in making sophisticated cardiac diagnostics accessible to millions of Indians who currently lack access to specialised heartcare facilities.
Published Aug 06, 2025 | 3:41 PM ⚊ Updated Aug 06, 2025 | 3:41 PM
The tool was developed jointly by Narayana Health's Clinical Research team and Medha AI.
Synopsis: The tool is now undergoing rigorous validation at multiple prestigious sites, including the Narayana Institute of Cardiac Sciences, Mazumdar Shaw Medical Center in Narayana Health City, select clinics across Bengaluru, and the Rabindranath Tagore International Institute of Cardiac Sciences in Kolkata. This multisite validation ensures the technology works effectively across diverse patient populations and clinical settings.
In a medical breakthrough that could transform cardiac care delivery in India, Bengaluru-based Narayana Health has developed the country’s first artificial intelligence model capable of predicting heart function from a standard ECG image in just 10 seconds.
The groundbreaking AI tool, developed jointly by Narayana Health’s Clinical Research team and Medha AI—the healthcare giant’s Advanced Analytics & AI division—represents a quantum leap in making sophisticated cardiac diagnostics accessible to millions of Indians who currently lack access to specialised heartcare facilities.
The innovation arrives at a critical juncture for India’s healthcare system. Heart failure affects nearly 10 million people across the country and accounts for a staggering 1.8 million hospital admissions annually. Yet access to echocardiography—the gold standard for evaluating heart function—remains frustratingly uneven, particularly in rural areas where specialist cardiac facilities are scarce or non-existent.
“This breakthrough represents much more than an AI model — it is a reflection of what Narayana Health stands for: combining clinical expertise, deep research, and cutting-edge technology to solve real problems at scale,” said Dr Devi Shetty, Founder and Chairman of Narayana Health.
“Detecting heart failure through a simple ECG in near real-time is a leap forward in making advanced diagnostics not just efficient but truly accessible to the masses,” he said.
The AI model functions like a highly trained cardiologist’s intuition, rapidly analysing ECG patterns to predict left ventricular ejection fraction (EF)—a critical measure of heart function that typically requires expensive echocardiogram equipment and specialized interpretation.
The development process was exhaustive and methodical. Trained on over 100,000 ECG images meticulously paired with corresponding echocardiogram reports, the AI model underwent rigorous testing that would make any pharmaceutical trial proud.
Dr Devi Shetty, Founder and Chairman of Narayana Health.
In external validation across 14 tertiary care centers involving more than 57,000 patients, the results were remarkable: the system identified 97% of individuals with severely reduced ejection fraction (≤35%).
Perhaps even more significantly, the AI flagged these high-risk patients an average of 58 days earlier than when echocardiography was eventually ordered through traditional clinical pathways—providing invaluable lead time for potentially life-saving interventions.
“What makes this innovation truly novel is its simplicity — our AI model can accurately assess heart function using just an image of an ECG, not complex signal data,” explained Dr Pradeep Narayan, Chief Scientific Officer at Narayana Health and Lead Researcher of the project. “Even a mobile phone photo of an ECG strip could be enough to generate a reliable assessment.”
This simplicity is revolutionary. Traditional cardiac assessment requires expensive equipment, specialized training, and often lengthy wait times. The new AI model transforms any basic ECG machine—already widely available in primary health centers across India—into a powerful cardiac screening tool.
The practical implementation of this technology demonstrates Narayana Health’s commitment to translating research into everyday clinical practice. The Medha AI model has been seamlessly integrated into Athma, Narayana Health’s proprietary Electronic Medical Record system, ensuring that the technology doesn’t disrupt existing clinical workflows but enhances them.
Currently, the system is undergoing rigorous validation at multiple prestigious sites, including the Narayana Institute of Cardiac Sciences, Mazumdar Shaw Medical Center in Narayana Health City, select clinics across Bengaluru, and the Rabindranath Tagore International Institute of Cardiac Sciences in Kolkata. This multisite validation ensures the technology works effectively across diverse patient populations and clinical settings.
“What sets Narayana Health apart is not just our ability to innovate, but to translate that innovation into everyday clinical practice,” said Dr Emmanuel Rupert, Managing Director and Group CEO of Narayana Health. “This AI model is a game-changer — it equips frontline clinicians with diagnostic insights that were previously only accessible through advanced infrastructure.”
The organisation’s vision extends far beyond hospital walls. Narayana Health plans to deploy this AI tool through its network of mobile medical units, specifically targeting underserved geographical areas where cardiac specialists are virtually non-existent. This approach could transform rural healthcare delivery, bringing specialist-level cardiac screening to remote villages and small towns.
“This innovation represents the true democratization of heart failure diagnosis,” emphasised Dr Deepak Padmanabhan, Senior Consultant and Strategic Lead for Cardiac EP Collegium at Narayana Health. “One of the biggest barriers to treatment is simply that heart failure often goes undiagnosed. Once identified, clinical pathways are relatively well established. But diagnosis itself can be complex and depends heavily on clinical experience.”
The AI tool addresses this diagnostic bottleneck by providing reliable insights even in settings with limited cardiac expertise. Its high accuracy, particularly in ruling out heart failure when the algorithm indicates normal function, makes it an invaluable decision-support tool for managing patients across diverse healthcare settings.
Dr PM Uthappa, Group Chief Medical Director at Narayana Health, stressed the importance of ethical AI implementation: “Responsible AI in healthcare means more than just innovation — it demands fairness in access, accountability in outcomes, transparency in how decisions are made, and explainable.”
The platform’s architecture enables future expansion beyond heart failure detection. Research teams are already working to broaden the AI’s capabilities to include detection of other echocardiographic abnormalities and early signs of coronary artery disease, potentially creating a comprehensive cardiac screening platform.
The innovation has garnered international attention, being presented at leading medical conferences including the Heart Rhythm Society Conference in Boston, APHRS in Sydney, IHRSCON in Kolkata, and HRX in Atlanta. The work has been published in the prestigious Indian Heart Journal and received the BMJ South Asia Award for Digital Innovation.
(Edited by Majnu Babu).
