--- title: "Sudden Cardiac Arrests Identified with New Artificial Intelligence System" url: https://noti.group/sudden-cardiac-arrests-identified-with-new-artificial-intelligence/ language: en publisher: "Noti Group" section: "Health" published: 2026-11-15T01:05:10.000Z updated: 2026-11-15T08:48:16.904Z id: ecce3b05-b2ea-40b7-a004-bdab2df9d566 source: "ScienceDaily Health https://www.sciencedaily.com/releases/2026/11/261112230456.htm" attribution: "Link to https://noti.group/sudden-cardiac-arrests-identified-with-new-artificial-intelligence/ and name Noti Group when you quote or summarize this story." --- # Sudden Cardiac Arrests Identified with New Artificial Intelligence System Researchers at UC Berkeley have developed an artificial intelligence system that can identify people at risk of sudden cardiac arrest using routine heart scans. More than 300,000 people in the United States die each year from sudden cardiac arrest, a condition where the heart's electrical system fails, causing it to stop pumping blood. This can happen to older adults with known heart conditions or seemingly healthy young athletes. Despite implanted defibrillators being able to prevent some of these deaths, doctors have long struggled to determine which patients need them most. A recent breakthrough by researchers at UC Berkeley has identified a previously overlooked warning signal in routine heart scans that could help identify people at risk before a fatal event occurs. This discovery was made possible using artificial intelligence trained on over 440,000 electrocardiograms collected from Sweden. The AI system compared the electrical patterns recorded in these scans with information from death certificates to teach itself subtle differences between healthy individuals, patients with known heart risks, and those who eventually died from sudden cardiac causes. By doing so, it was able to recognize unique patterns that were previously overlooked. The researchers then tested their technology using thousands of additional patient records from the United States and Taiwan, evaluating whether the patterns could reliably identify people at heightened risk in different populations. This step was crucial in determining whether the AI system's findings would be applicable globally. Current screening methods focus on how much blood the heart pumps with each contraction to assess whether someone might benefit from an implanted defibrillator. However, this measurement only identifies a group of patients with an annual sudden cardiac death rate of 4.6%. In contrast, the AI system identified a group with an annual death rate of 7%, showing its greater ability to distinguish patients facing serious danger. The results suggest that AI can identify vulnerable patients more effectively than conventional tests doctors typically rely on. This has significant implications for how heart health is assessed and treated in the future. The AI system has been able to identify individuals at elevated risk of heart problems, including thousands who may not have been considered high-risk under conventional medical guidelines. These individuals would typically be classified as low-risk using standard methods, highlighting a significant shortcoming in current heart health assessment and treatment strategies. A major advantage of the AI technology is its ability to work with ordinary EKG images, which are already widely available in hospitals and medical centers around the world. This means that the system could potentially improve screening without requiring an entirely new type of medical test, making it a more practical solution for widespread implementation. The findings have significant implications for how heart health is assessed and treated, particularly when it comes to determining who should receive an implanted defibrillator. Physicians may soon be able to make more informed decisions about which patients are at risk of sudden cardiac death, thanks to the AI system's ability to identify subtle electrical patterns in the heart. Sudden cardiac death remains one of medicine's most puzzling and deadly mysteries. This is according to Obermeyer, an emergency physician whose research combines artificial intelligence, medicine, and health policy. The suddenness of these deaths often leaves researchers without a window into what happens inside the heart immediately before it stops. Autopsies may reveal structural abnormalities such as blocked blood vessels or hardened tissue, but they cannot provide a complete picture of how the heart was functioning in the moments leading up to death. This lack of insight makes it challenging for doctors to identify those at risk and intervene early. Obermeyer notes that one of the most frustrating aspects of sudden cardiac death is its preventability. "We have the cure," he said, but we can't figure out who needs it. If patients knew they were at high risk, they would likely seek medical attention and undergo procedures to mitigate their danger. Currently, doctors rely on a measurement of heart function known as ejection fraction, which gauges how effectively the heart pushes blood with each beat. Patients whose hearts pump below a certain threshold may be considered candidates for an implantable cardioverter defibrillator. This device can deliver an electrical shock when a life-threatening rhythm develops. However, obtaining this measurement often requires extensive medical evaluation and testing. Many people who eventually experience sudden cardiac arrest never receive such testing because they lack obvious risk factors. As a result, potentially lifesaving interventions may be delayed or missed entirely. The approach also has limitations for patients identified as high-risk. A significant proportion of those with implanted defibrillators never require the device to deliver a shock. Consequently, many individuals undergo expensive and invasive procedures for emergencies that may not occur. Thousands of people die unexpectedly each year without being recognized as high-risk patients. According to researchers, many of these deaths could have been prevented if timely interventions had been possible, but current methods often fall short. The identification of more accurate warning signs is seen as a crucial step towards directing preventive treatments towards the most vulnerable patients. To develop an algorithm capable of recognizing deadly heart patterns, researchers turned to large-scale medical data collections from around the world. They began with six years' worth of EKG recordings from Sweden's unified healthcare system, linking these scans to death certificates to examine the electrical patterns associated with sudden cardiac death. This allowed them to compare the recordings of patients who experienced sudden cardiac death with those of other patients, providing valuable insights into potential warning signs. The massive undertaking of gathering and organizing the data required for this project was a significant challenge in itself. Obermeyer estimates that assembling the datasets took around 10 years to complete, and involved contributions from two organizations he helped establish: Dandelion Health and Nightingale Open Science. Both organizations aim to advance medical research by improving access to data and leveraging artificial intelligence. As Obermeyer noted, having a team at his disposal has been beneficial in pursuing projects that may not yield immediate results but could ultimately lead to significant breakthroughs. He emphasized the value of working within an academic environment like Berkeley's. The researchers are now exploring practical applications for their system in real-world healthcare settings. They have initiated collaborations with medical systems in Sweden, Taiwan, and the United States to integrate the algorithm into hospital EKG databases. Physicians would use the system to review existing heart scans for signs of elevated risk and identify patients who might benefit from closer monitoring. Those patients could then be contacted by their doctors about the possibility of wearing a patch that continuously records their heart's electrical activity. These longer recordings could provide valuable insights into the biological processes underlying the warning signals detected by the AI, potentially shedding light on why certain hearts are more susceptible to fatal electrical disturbances. The groundbreaking discovery of a hidden heart signal by an AI algorithm has significant implications for medical research and treatment. Researchers are now able to uncover patterns in heart activity that were previously unknown, offering new avenues for understanding the electrical disturbances responsible for sudden cardiac death. This could lead to improved prevention strategies and more effective treatments for patients at risk. The use of artificial intelligence in this study highlights its potential to shed light on medical mysteries that have resisted conventional approaches. By analyzing complex data sets, AI can identify subtle patterns and connections that may elude human observers. As researchers continue to explore the insights provided by these longer recordings, they may uncover new biological processes underlying warning signals detected by the AI. This could lead to a deeper understanding of why certain hearts are more susceptible to fatal electrical disturbances. Ultimately, this innovative application of AI in medical research has the potential to revolutionize our approach to healthcare and significantly improve patient outcomes. --- Source: [ScienceDaily Health](https://www.sciencedaily.com/releases/2026/11/261112230456.htm) Published by Noti Group: https://noti.group/sudden-cardiac-arrests-identified-with-new-artificial-intelligence/