[Constructing a cataplexy face prediction model for narcolepsy type 1 based on ResNet-18]

Y Q Yuan1, Y Zhu2, B Z Chen3

  • 1Department of Neurology, the Second Affiliated Hospital, Jiangxi Medical College, Nanchang University, Nanchang 330006, China.

PubMed
Summary

A deep learning model using ResNet-18 accurately identifies cataplexy facial features in clinical videos. This AI tool shows high sensitivity and specificity for diagnosing cataplexy in narcolepsy type 1 patients.