Search

Objective. In this paper, a new approach of extracting and measuring the variability in electroencephalogram (EEG) was proposed to assess the depth of anesthesia (DOA) under general anesthesia. Approach. The EEG variability (EEGV) was extracted as a fluctuation in time interval that occurs between two local maxima of EEG. Eight parameters related to EEGV were measured in time and frequency domains, and compared with state-of-the-art DOA estimation parameters, including sample entropy, permutation entropy, median frequency and spectral edge frequency of EEG. The area under the receiver-operator characteristics curve (AUC) and Pearson correlation coefficient were used to validate its performance on 56 patients. Main results. Our proposed EEGV-derived parameters yield significant difference for discriminating between awake and anesthesia stages at a significance level of 0.05, as well as improvement in AUC and correlation coefficient on average, which surpasses the conventional features of EEG in detection accuracy of unconscious state and tracking the level of consciousness. Significance. To sum up, EEGV analysis provides a new perspective in quantifying EEG and corresponding parameters are powerful and promising for monitoring DOA under clinical situations.

Hypertension affects huge number of people around the world. It also has a great contribution to cardiovascular and renal related diseases. This study investigates the ability deep convolutional autoencoder (DCAE) to generate the continuous arterial blood pressure (ABP) by only utilizing the photoplethysmography (PPG) to generate the continuous ABP. The total of 18 patients is utilized. LeNet-5 and U-Net based DCAEs, respectively for LDCAE and UDCAE, are compared to the MP60 IntelliVue Patient Monitor, as the golden standard. Moreover, in order to investigate the data generalization, leave-one-out cross-validation (CV) method is conducted. The results show that the UDCAE provides superior results in producing the SBP estimation. Meanwhile, LDCAE gives a slightly better for the DBP prediction. Finally, the genetic algorithm (GA) based optimization deep convolutional autoencoder (GDCAE) is further administered to optimize the ensemble of the CV models. The results reveal that the GDCAE is superior to either the LDCAE or UDCAE. For conclusion, this study reveals that the SBP and DBP can also be accurately achieved by only utilizing the single PPG signal.

Convolutional neural network (CNN) have been widely used in various fields in recent years. However, the CNN method is rarely used in EEG studies to assess the depth of anesthesia (DOA) in patients. In this study, EEG signal is used as the input to the convolutional, long short-term memory, fully connected deep neural networks (CLDNN) to predict DOA using continuous wavelet transform (CWT). According to the bispectral (BIS) index and signal quality index (SQI) measured by medical equipment, the anesthesia state is divided into anesthesia light (AL), anesthesia OK (AO), anesthesia deep (AD). The computing window of CWT is 120s. Moreover, 75% overlapped computing window is set to enrich medical data. Through different models, the epoch, timestep and input size of the CWT image were changed to get the best experimental results: AL was 82%, AO was 89%, and AD was 87%. The overall accuracy of the model is 87.79%, and AL and AD can be fully predicted. Ministry of science and technology (MOST) of Taiwan

A heart rate variability (HRV) measuring system and its analysis method have been developed in this study. It is composed of a hardware measuring system based on a noninvasive photoplethysmographic (PPG) signal measuring device to acquire oxyhemoglobin saturation using pulse oximetry (SpO2) signals and a further software package including the methods used to filter and analyze the signals for heart rate variability. Firstly, an experiment is designed for measuring heartbeat using the system to observe whether the empirical mode decomposition(EMD) can really inhibit noise or not on one volunteer with 10 minutes repeated for 10 times. Then, the hardware system and analysis method are tested on another 10 volunteers before and after receiving cold face immersion. The results of the first experiment have no significant difference with commercial instrument (p > 0.05), but the results using EMD perform better when signals are contaminated by artifacts. The second part experiment is subdivided into two stages. The results show that HR values at each stage have no significant difference with commercial instrument (p > 0.05). The LF significantly decreases from 0.33 0.03 to 0.31 0.03, while HF significantly increases from 0.41 0.07 to 0.43 0.07 indicating cold face immersion can increase parasympathetic and decrease sympathetic actions. Hence, LF/HF changes significantly from high (0.85 0.17) to low (0.74 0.17) before and after adding stimulation. Due to the reasons above, it confirms that the developed system can measure heartbeat and observe the heart rate variability. So the findings of this research may be useful for developing a low-cost and a miniaturized pulse oxymeter system to continuously measure HR and HRV for the purpose of convenience, portability, and operability.

The definition of the depth of anesthesia (DOA) is still controversial and its measurement is not completely standardized in modern anesthesia. Power spectral analysis is an important method for features detection in electroencephalogram (EEG) signals. Several spectral parameters derived from EEG have been proposed to measuring depth of anesthesia (DOA) for clinical application. In present paper, an improved method based on phase-rectified signal averaging (PRSA) is designed to improve the predictive accuracy of relative alpha and beta power, a frequency band power ratio, total power, median frequency (MF), spectral edge frequency 95 (SEF95), and spectral entropy for assessing the anesthetic drug effects. Fifty six patients undergoing general anesthesia in operation room are studied. All EEG signals are continuously recorded from the awake state to the end of the recovery state and then filtered using multivariate empirical mode decomposition (MEMD). All parameters are evaluated using the commercial bispectral index (BIS) and expert assessment of conscious level (EACL), respectively. The ability to predict DOA is estimated using the area under the receiver-operator characteristics curve (AUC). All indicators based on improved method can clearly discriminate the conscious state from anesthetized state after filtration (p