Vuno, Vuno Med DeepCass™ first prospective study results published in an international academic journal
Wednesday, September 13, 2023
VunoMed DeepCass™, Vuno's AI-based cardiac arrest prediction medical device, proved clinical effectiveness for the first time through a prospective study using real world data (RWD) from clinical sites.
Medical artificial intelligence company Vuno (CEO Lee Ye-ha) published a multi-center clinical research paper that proved the clinical effectiveness of VunoMed DeepCath™ through the first prospective study in 'Critical Care (IF 19.344), the world's most authoritative academic journal in the field of critical care medicine. was published [i] on the 13th.
Prospective studies using actual data have high reliability among medical staff as the data is obtained from actual clinical sites where various variables exist. So far, most research on AI medical devices has been conducted as a retrospective study using medical data from previous periods. This study by Vuno Med is meaningful in that Vuno Med Deepcas™, which has previously proven its clinical effectiveness through several retrospective studies, showed consistent predictive performance and high reliability in actual clinical settings and prospective studies.
The Vuno research team conducted a multicenter study using data from patients hospitalized in general wards of four tertiary medical institutions with different sizes, locations, and medical environments. Participating medical institutions included Seoul National University Hospital, Seoul National University Bundang Hospital, Inha University Hospital, and Dong-A University Hospital. Each medical institution targeted a total of 55,083 patients hospitalized in general wards over a three-month period, using high-risk patient prediction systems such as the existing NEWS (National Early Warning Score), VunoMed Deepcas™ in-hospital cardiac arrest (IHCA), and unexpected intensive care unit anteroom. (UIT) Performance, including prediction accuracy and false alarm rate, was compared and analyzed.
As a result of the study, the prediction performance of VunoMed DeepCass™ was 0.869 based on AUROC , a performance indicator indicating prediction accuracy , which was superior to existing methods (NEWS 0.767, MEWS 0.756). Compared to the same sensitivity, the number of alarms per 1,000 beds was reduced by more than half, and the rate of high-risk alarms leading to actual medical staff action was the highest compared to existing methods, proving the alarm's high reliability. In addition, it was shown to be effective without any restrictions such as the patient's age, gender, or time of occurrence.
Through this study, Vunomed DeepCass™ demonstrated excellent predictive performance and versatility for high-risk patients in the first prospective study. In addition, Vuno showed consistent performance with the results of retrospective studies of the product that have been proven through various world-renowned academic journals. This is expected to be additional evidence supporting the fact that VunoMed DeepCass™, which is currently spreading rapidly in medical settings, can dramatically improve the safety management of hospitalized patients in actual medical environments.
Lee Ye-ha, CEO of Vuno, said, “The results of this study are significant in that they demonstrated the consistent predictive performance and versatility of Vuno Med DeepCass™ by successfully conducting the first multi-center prospective study with various variables and high difficulty.” “We will conduct additional prospective research in the future to continue to increase the reliability of the product in clinical settings, while also doing our best to ensure patient safety by introducing it to more medical settings as essential medical care,” he said.
Source: vuno.co