基于 UKF 的脑部麻醉深度状态监测

根据脑电图信号进行脑状态跟踪是一个具有挑战性的问题。基于 NMM 的方法可以在神经科学及临床应用中推测潜在生理学变化，同时跟踪脑状态，具有广阔的应用前景。NMM 中的 IPSP 振幅和 IPSP 速率常数可以直观反映受试者麻醉引起的生理学变化。提出采用无迹卡尔曼滤波（UKF）估计神经群模型参数，并讨论所估计的参数随麻醉深度的变化情况。结果进一步说明 IPSP 振幅参数估计和 PSP 速率常数会因受试者不同而不同，但所有参数会随时间即麻醉水平的提升而增加。IPSP 振幅估计值显示每个受试者的最大波动，反映了对瞬时脑电图波动的敏感性。

Brain Anesthesia Depth State Monitoring Based on UKF

Monitoring brain states based on EEG signals is challenging. The NMM-based approach has shown promise in predicting potential physiological changes and tracking brain states in both neuroscience and clinical applications. The IPSP amplitude and IPSP rate constant in NMM can directly reflect the physiological changes caused by anesthesia. We propose the method to estimate the parameters of the neural mass model by unscented Kalman filter and discuss the variation of the parameters with the depth of anesthesia. The results further illustrate that the estimated value of IPSP amplitude parameter and IPSP rate constants vary from subject to subject，but all parameters increase over time，that is，as the level of anesthesia increases. The IPSP amplitude estimate shows the maximum fluctuation for each subject，reflecting the sensitivity to transient EEG fluctuations.