This paper proposes a new method to quantify the strength and direction of the information flow, that is, Time-delay multiscale symbolic phase compensated transfer entropy (denoted as $SPcTE(\tau ,s)$). We integrate time delay factors and time scale factors into the study of transfer entropy and use compensated transfer entropy to estimate transfer entropy to effectively eliminate the influence of instantaneous effects. Variational modal decomposition and wavelet analysis are used to preprocess experimental data to reduce noise. Artificial data experiments show that our method can capture the direction and size of the information flow between two time-delayed time series, locate the time point of the information transmission, and highlight the interaction between the two series on different time scales. In the study of the $SPcTE(\tau ,s)$ of the Cyclic Alternating Pattern (CAP) in the clinical cases of adult sleep-disordered breathing (SDB), sleep disorders such as REM behavior disorder (RBD), bruxism (BRUX), insomnia (ins) and other states, it is found that they have different performance on different time delay factors and time scales. This method provides a more comprehensive research framework and provides important information in the potentially dynamic process.

本文提出了一种量化信息流强度和方向的新方法，即时延多尺度符号相位补偿传递熵（表示为 $\mathrm{小号}PCŤE（\tau ，s）$）。我们将时延因子和时标因子整合到转移熵的研究中，并使用补偿转移熵来估计转移熵，以有效消除瞬时效应的影响。变分模态分解和小波分析用于预处理实验数据以减少噪声。人工数据实验表明，该方法可以捕获两个时延时间序列之间信息流的方向和大小，确定信息传输的时间点，并突出显示两个时间序列在不同时间尺度上的相互作用。在研究中$\mathrm{小号}PCŤE（\tau ，s）$发现成人睡眠呼吸障碍（SDB），睡眠障碍（例如REM行为障碍（RBD），磨牙症（BRUX），失眠（ins）和其他状态的临床病例中的循环交替模式（CAP）的发现它们在不同的时间延迟因子和时间尺度上具有不同的性能。这种方法提供了更全面的研究框架，并在潜在的动态过程中提供了重要的信息。