Key Takeaways
- Researchers have created an AI‑powered tool that analyses a standard electrocardiogram (ECG) in under two seconds to flag likely heart failure or heart valve disease.
- The model was trained on millions of patient records and can detect up to 81 % of heart‑failure cases and up to 90 % of valve‑disease cases in a large U.S. trial.
- While the AI cannot replace a definitive echocardiogram, it serves as a rapid triage system to prioritize patients for further ultrasound scanning, potentially cutting months‑long waits.
- Experts stress that earlier detection enabled by the tool can start life‑saving treatment sooner, improving outcomes for millions.
- The technology could be deployed opportunistically on all hospital ECGs or adapted into handheld devices for point‑of‑care use.
A New “Superhuman” AI for Heart‑Disease Screening
Doctors have unveiled an artificial‑intelligence system capable of spotting signs of heart failure and heart valve disease from a routine electrocardiogram (ECG) in “the blink of an eye.” The tool, described by its creators as a “superhuman AI,” extracts far more information from the ECG trace than a human clinician can typically discern, turning a century‑old bedside test into a powerful early‑warning screen.
What a Traditional ECG Does—and What It Misses
A standard ECG records the heart’s electrical activity, revealing rate and rhythm abnormalities that have guided the diagnosis of heart attacks and arrhythmias for over a hundred years. However, the test alone cannot reveal structural problems such as weakened heart muscle or leaky valves; those conditions normally require an echocardiogram, an ultrasound scan that patients often wait months to obtain.
How the AI Works in Real Time
By analysing millions of ECG recordings linked to confirmed diagnoses, the AI model learns subtle patterns indicative of myocardial dysfunction or valvular pathology. In practice, the algorithm processes a fresh ECG and returns a risk score in less than two seconds, effectively flagging patients who are “highly likely” to have heart failure or valve disease.
Unveiling the Breakthrough at Europe’s Largest Heart Conference
The development was presented to thousands of delegates at the European Society of Cardiology (ESC) annual congress in Munich, the world’s largest gathering of cardiovascular specialists. The venue underscored the potential global impact of a technology that could be applied wherever ECGs are routinely performed.
Why Early Detection Matters
Early identification of heart failure and valve disease is crucial because it allows clinicians to initiate guideline‑directed therapies before patients become dangerously unwell. As Dr Sonya Babu‑Narayan, a consultant cardiologist and clinical director of the British Heart Foundation (BHF), noted, “It is exciting to see that AI can now deliver a read‑out from an ECG in what feels like the blink of an eye.” She added that such technology “could be a solution to help fast‑track the patients who are most likely to have a heart abnormality. When it comes to the heart, earlier diagnosis and treatment saves and improves lives.”
Trial Performance: Sensitivity and Specificity
In a U.S. trial involving 67,000 patients, the AI tool identified up to 81 % of individuals who actually had heart failure and up to 90 % of those with heart valve disease. These figures represent the sensitivity of the model—its ability to catch true positives—while maintaining a low false‑positive rate that makes it suitable for triage rather than definitive diagnosis.
The AI as a Triage Tool, Not a Stand‑Alone Test
Researchers are clear that the algorithm cannot on its own confirm or exclude heart failure or valve disease. Instead, it provides a “very strong indication” that a patient may be affected, prompting rapid referral for an echocardiogram. Prof Fu Siong Ng, a professor of cardiology at Imperial College London, explained, “Patients can often wait several months for a heart ultrasound scan after being referred for one by their doctor. This makes it exciting that our technology could identify patients most at risk of heart failure and heart valve disease, so they could be prioritised for scans faster and more urgently.”
Opportunistic Screening and Future Hardware
Beyond flagging suspected cases, the AI could be run opportunistically on every ECG performed in a hospital, uncovering hidden disease in patients who underwent the test for unrelated reasons. Dr Ahmed El‑Medany, a BHF clinical research fellow who led the Imperial analysis, envisions the next step: “The next challenge would be to design handheld AI‑led ECG readers for healthcare professionals to use.” Such devices could bring the screening power to clinics, emergency departments, or even community health settings.
Broader AI Applications Highlighted at the Congress
The same ESC session featured other AI innovations, including research from the University of Tokyo and the Institute of Science Tokyo showing that five‑second facial‑video analysis can detect undiagnosed high blood pressure and type 2 diabetes with high accuracy. Like the ECG‑AI, these tools aim to catch silent conditions early, reducing the burden of untreated disease on individuals and health systems.
Conclusion: A Step Toward Faster, More Equitable Cardiac Care
By marrying the ubiquity of the ECG with the pattern‑recognition strength of deep learning, the new AI system promises to shrink diagnostic intervals from months to moments. While it will not replace the echocardiogram, its role as a rapid triage mechanism could accelerate treatment initiation, alleviate specialist workload, and ultimately save lives—especially in settings where access to advanced imaging is limited. As the technology moves from research labs to handheld devices, the vision of “superhuman” cardiac screening inches closer to everyday clinical reality.
https://www.theguardian.com/technology/2026/aug/31/superhuman-ai-tool-spots-heart-disease