AI Predicts Cardiac Arrest Risk in Hospitalized Patients... ECG Monitoring at Home, Too

'VUNO Med-DeepCARS' Builds Clinical Evidence at 150 Hospitals
Home ECG Monitor 'HATIV' Enables Early Detection of Atrial Fibrillation
Wearable Devices Expand Medical AI to Homes and Communities

The use of medical AI is expanding beyond hospitals. It can predict the risk of cardiac arrest in hospitalized patients and help detect arrhythmia early at home through simple electrocardiogram measurements.

Joo Seonghun, Chief Technology Officer (CTO) of VUNO, introduces the company's AI solutions currently used in clinical settings and their future development at the “Academic Event Celebrating the 70th Anniversary of Eulji University Medical Center,” held at The Grand Lotte Seoul in Jung-gu, Seoul, on the 30th. His presentation was titled “Expanding the Use of Medical AI from Hospitals to Homes.”

Joo Seonghun, Chief Technology Officer (CTO) of VUNO, introduced AI solutions currently used in clinical settings and their future development at the Academic Event Celebrating the 70th Anniversary of Eulji University Medical Center on October 30. His presentation was titled “Expanding the Use of Medical AI from Hospitals to Homes.”

Joo identified as a key challenge for medical AI not only the accuracy of the technology but also demonstrating its effects on patients in actual clinical settings. “It is important for medical AI to make accurate predictions, but it is also important to show how clinically effective the results are when it is actually introduced in hospitals,” he stressed.

One notable example is VUNO’s AI-based cardiac arrest risk monitoring solution, VUNO Med-DeepCARS. In general wards, nurses typically measure patients’ vital signs, including blood pressure, pulse, respiration and temperature, at eight-hour intervals. The solution analyzes these records using an AI algorithm and calculates the risk of cardiac arrest on a scale of 0 to 100. A higher score indicates a greater risk. Medical staff can view the risk levels of ward patients at a glance and prioritize monitoring of patients whose scores have risen sharply. The system is typically set to issue a high-risk alert when a score reaches 90 or above.

At one hospital, a patient whose DeepCARS score rose sharply was closely monitored, transferred to the intensive care unit for early intervention and ultimately discharged. “AI is being used to identify which patients among many need to be monitored most closely,” Joo explained.

Clinical evidence is also accumulating. DeepCARS, which was approved by the Ministry of Food and Drug Safety in 2021, has built up clinical evidence over approximately four years through the deferred evaluation system. It is now being used at about 150 hospitals in South Korea and is set to undergo a New Health Technology Assessment this year. Research has also found significant reductions in cardiac arrest and mortality rates at hospitals that adopted DeepCARS.

Joo particularly emphasized that “medical AI should complement, not replace, the existing work systems of medical staff.” DeepCARS, too, is designed not to replace a hospital’s entire rapid response system, but to detect early signs that a patient’s condition may deteriorate.

As a technology for use outside hospitals, Joo introduced HATIV, a home-use electrocardiogram device. When the user touches the device with both hands and a foot, it measures an electrocardiogram for 30 seconds. AI then analyzes the results and indicates whether the user may have atrial fibrillation, among other conditions. The results also display key indicators such as heart rate, QRS interval and QT interval.

Atrial fibrillation is an arrhythmia in which the heart beats irregularly. It can increase the risk of blood clots and stroke, making early detection and treatment important. However, because it occurs intermittently, an abnormality may not appear when an electrocardiogram is taken at a hospital. HATIV was developed to let patients take electrocardiogram readings at home whenever needed and keep a record of them. At a public health center in a village on an island in Sinan, South Jeolla Province, the device was used to identify an elderly patient whose electrocardiogram had appeared normal but who experienced repeated episodes of atrial fibrillation over approximately six hours. The patient was immediately referred to a hospital and received prompt treatment.

At the Busan International Medical Equipment & Hospital Equipment Exhibition (KIMES BUSAN 2026), held at BEXCO in Busan from the 23rd to the 25th, VUNO is operating a self-service ECG experience zone featuring key products, including the kiosk-style HATIV K30 electrocardiogram device.

However, as the use of home devices expands, it is also important for people to understand the results correctly. Even a temporary abnormal signal can cause anxiety for patients. “Premature contractions can occur temporarily, including the day after drinking coffee or alcohol,” Joo said. “Explaining the results accurately to patients will be a challenge going forward.”

VUNO also plans to develop technology that uses wearable devices to continuously measure the vital signs of hospitalized patients and link the data to AI-based patient risk monitoring. The aim is to supplement the blood pressure, respiratory rate and temperature measurements that nurses currently take at regular intervals, reducing gaps in monitoring. Further development could make it possible to continuously monitor discharged patients’ health at home and connect them with medical institutions when their vital signs show signs of risk.

“VUNO’s goal is to make medical technologies used in large hospitals available at primary care institutions, in local communities and at home,” Joo said. “We will continue to build clinical evidence so that medical AI can bring meaningful change to patients in real-world settings.”

This content was produced with the assistance of AI translation services.