VUNO Med®-DeepCARS™ Receives Green Light from MFDS for Conducting Clinical Trials
- 06. 30. 2020
Deep Learning-based Cardiac Arrest Prediction SW, VUNO Med®-DeepCARS™ Receives Green Light from MFDS for Conducting Clinical Trials
VUNO Inc., South Korean artificial intelligence (AI) developer announced that the clinical trial plan for its deep learning-based cardiac arrest prediction software, VUNO Med®–DeepCARS™ received approval from the Ministry of Food and Drug Safety. This is the first biosignal-based AI medical device to take steps towards commercialization in the Korean market, where research was largely based on image analysis in radiology and pathology.
VUNO Med®–DeepCARS™ is a software that analyzes vital signs such as heart rate, respiratory rate, systolic blood pressure, diastolic blood pressure, and body temperature, recorded in electronic medical records (EMR) of patients in the general ward, and provides information on the likelihood of cardiac arrest. It predicts the level of risk within the next 24 hours with corresponding confidence scores to assist healthcare professionals in taking preemptive measures as well as make accurate diagnostic decisions.
Inpatients often begin to manifest abnormal symptoms a few hours prior to the occurrence of cardiac arrest, enabling rapid response when it is detected early. However, timely detection of such symptoms and making instant decisions taking into account all vital signs in hospitals without an accurate scoring system are restricted. Moreover, conventional methods such as MEWS (Modified Early Warning Score) measure risk scores at a certain point of time by aggregating scores for each vital sign by predefined rules, so they are known to have low sensitivity and high false alarm rates leading to alarm fatigue.
VUNO Med®–DeepCARS™ applied the latest deep learning technique, the Recurrent Neural Network (RNN) that analyzes the patients’ vital signs recorded from the moment of hospitalization to improve accuracy compared to the existing score systems. According to a research paper published in Critical Care Medicine (CCM), VUNO Med®–DeepCARS™ was twice as sensitive as MEWS for the same number of alarms, and the number of alarms was almost half (59.6%) for the same sensitivity, proving its low false alarm rate.
If VUNO Med®–DeepCARS™ can be implemented in the clinical field, healthcare providers will be able to effectively predict in-hospital cardiac arrest and take preemptive measures accordingly. Research has shown that the number of cardiac arrest and death rates were reduced with the implementation of the conventional score system , therefore, it is expected that the use of the far superior product, VUNO Med®–DeepCARS™ can bring about a much more meaningful drop in both numbers.
“We have high hopes for VUNO Med®–DeepCARS™, as we have been working on it for a long time, and is one of the most prominent research results in the non-medical imaging field,” said Hyun-Jun Kim, CEO of VUNO, and added that, “We will do our very best to save more lives by putting this product to more use in the clinical field.”
VUNO succeeded in commercializing a wider array of AI medical solutions used to analyze various types of modalities images generated in all stages of medical practice from diagnosis to treatment radiology, pathology, vital signs, and voice recognition. VUNO launched various AI solutions at home and abroad including ▷VUNO Med®-BoneAge™, the first ever AI medical device in Korea, ▷VUNO Med®-DeepBrain™, ▷VUNO Med®-Chest X-ray™, ▷VUNO Med®-LungCT AI™, ▷VUNO Med®-Fundus AI™, and ▷VUNO Med®-DeepASR™, of which five recently received the CE mark.
Yerim Kim / PR Manager, VUNO lnc.