jibi to rinsho
Online ISSN : 2185-1034
Print ISSN : 0447-7227
ISSN-L : 0447-7227
Original Article
A Practical Influenza Diagnosis Using Artificial Intelligence-equipped Medical Devices
Atsushi KAWANOShogo OYAMADASaeko MATSUZAKIKyoko ITOIkumi KUBOYAEmiri KOREYASUTomomi SAITORisa TAKASUGIAyumi MIYASAKANaoko WAKASAHideki TAKEMURA
Author information
JOURNAL FREE ACCESS

2023 Volume 69 Issue 6 Pages 401-414

Details
Abstract

Influenza testing was performed using an artificial intelligence (AI)-equipped medical device (market name:Nodoca). The concordance rate between this test and the conventional rapid test was 75.4%. In the relatively early stage from the onset of illness to testing, both cases that were negative according to the conventional rapid test due to insufficient viral growth as in the previous report and were "detected" by this test was 9.6%. In contrast, in 15.1% of cases, the rapid test was positive, but the device failed to detect the virus, suggested that improving the accuracy of AI-equipped medical devices is essential to ensure their widespread use in clinical practice. Most operators (80-90%) had favorable opinions, noting that testing with this device was easier than conventional testing, leading to better understanding; however, opinions were divided regarding influenza types (A and B). An examination with this device was relatively easy, with suffocation only noted in 15.3% and pain in 10.5%, but there were three cases in which the tongue was stiff, and pharyngeal findings could not be obtained. This device seemed to be effective in early-onset cases (especially within 18 h) and in those who refused nasopharyngeal wipes.

Content from these authors
© 2023 JIBI TO RINSHO KAI
Next article
feedback
Top