Research on G-induced loss of consciousness early warning model based on cerebral oxygen saturation
CAO Zhengtao
JIN Zhao
WANG Cong
TIAN Zhen
YANG Minghao
Abstract：Objective To provide early warning for pilots'G-induced loss of consciousness(G-LOC)status under different acceleration exposure modes based on cerebral oxygen saturation.Methods This study selected 33 volunteers for manned centrifuge training,applied different acceleration exposure modes,and measured physiological indicators such as cerebral oxygen saturation.An extreme learning machine(ELM)model was established to predict G-LOC status.Results The established G-LOC early warning model performed well on experimental data trained on manned centrifuges,with an average accuracy of 93.61％,an average G-mean of 0.963 7,a true positive rate of 99.45％,and a true negative rate of 93.51％.Conclusion The ELM model based on cerebral oxygen saturation established in this study can effectively predict the critical syncope state of G-LOC.In the future,the construction of G-LOC early warning mechanism based on cerebral oxygen saturation can be further improved by expanding the sample data size and combining flight training with combat practice.
Keywords：accelerationcerebral oxygen saturationG-induced loss of consciousness early warning algorithmmachine learning
Publication Date：2025-01-27
Online Publishing Date：2026-08-26(First online date of this platform, not the publication date of the document)