South Korean machine learning developer VUNO has received regulatory approval from the country’s medical device regulator for its AI medical device that can read vital signs collected from patient monitoring devices and predict if a patient will suffer a cardiac arrest.

South Korea’s Ministry of Food and Drug Safety (MFDS) has approved the VUNO Med–DeepCARS, an AI medical device for cardiac arrest prediction through four primary vital signs. These comprise: blood pressure (diastolic and systolic); respiratory rate; heart rate; and body temperature.

The Vuno-DeepCARS device collects these data points from a hospital electronic medical record system, which gathers them online from connected patient monitors.

According to VUNO, its device has been shown to provide reliable early cardiac arrest prediction which can enable medical practitioners to mount a rapid and effective response.

VUNO Med–DeepCARS has been studied in a clinical trial at the Asan Medical Centre in Seoul, and has articles published in emergency medicine journals including; Resuscitation, Journal of the American Heart Association (JAHA) and Critical Care Medicine (CCM).

General hospital wards have been noted for having limited capacity to respond to rapidly-deteriorating patients compared to intensive care units due to the difficulty of constant patient monitoring. In-hospital cardiac arrest, with a high mortality rate of 75%, strikes over 290,000 hospital patients each year in the United States alone.

Once implemented in clinical practice, VUNO Med–DeepCARS could enable more rapid and efficient responses to in-hospital cardiac arrest by predicting a cardiac event using vital signs automatically collected from a patient’s EMR. The ability to utilise key vital signs, which are routinely collected from in-ward patients, can support broader adoption of the device in a range of clinical environments.

VUNO Chairman Lee Yeha said, “We seek to help save lives by urging the rapid adoption of VUNO Med–DeepCARS in hospitals,” adding that “the regulatory approval of VUNO Med–DeepCARS heralds the beginning of a wider application of our highly-promising biosignal-based AI technology.”