Robust PPG-Based Mental Workload Assessment System Using Wearable Devices.

Heart rate variability (HRV) has been used in assessing mental workload (MW) level. Compared with ECG, photoplethysmogram (PPG) provides convenient in assessing MW with wearable devices, which is more suitable for daily usage. However, PPG collected by smartwatches are prone to suffer from artifacts. Those signal corruptions cause invalid Inter-beat Intervals (IBI), making it challenging to evaluate the HRV feature. Hence, the PPG-based MW assessment system is difficult to obtain a sustainable and reliable assessment of MW. In this paper, we propose a pre- and post- processing technique, called outlier removal and uncertainty estimation, respectively, to reduce the negative influences of invalid IBIs. The proposed method helps to acquire accurate HRV features and evaluate the reliability of incoming IBIs, rejecting possibly misclassified data. We verified our approach in two open datasets, which are CLAS and MAUS. Experiment results show proposed method achieved higher accuracy (66.7% v.s. 74.2%) and lower variance (11.3% v.s. 10.8%) among users, which has comparable performance to an ECG-based MW system.

Noise-Level Aware Compressed Analysis Framework for Robust Electrocardiography Signal Monitoring.

Compressed sensing (CS) has drawn much attention in electrocardiography (ECG) signal monitoring for its effectiveness in reducing the transmission power of wireless sensor systems. Compressed analysis (CA) is an improved methodology to further elevate the system's efficiency by directly performing classification on the compressed data at the back-end of the monitoring system. However, conventional CA lacks of considering the effect of noise, which is an essential issue in practical applications. In this work, we observe that noise causes an accuracy drop in the previous CA framework, thus discovering that different signal-to-noise ratios (SNRs) require different sizes of CA models. We propose a two-stage noise-level aware compressed analysis framework. First, we apply the singular value decomposition to estimate the noise level in the compressed domain by projecting the received signal into the null space of the compressed ECG signal. A transfer-learning-aided algorithm is proposed to reduce the long-training-time drawback. Second, we select the optimal CA model dynamically based on the estimated SNR. The CA model will use a predictive dictionary to extract features from the ECG signal, and then imposes a linear classifier for classification. A weight-sharing training mechanism is proposed to enable parameter sharing among the pre-trained models, thus significantly reducing storage overhead. Lastly, we validate our framework on the atrial fibrillation ECG signal detection on the NTUH and MIT-BIH datasets. We show improvement in the accuracy of 6.4% and 7.7% in the low SNR condition over the state-of-the-art CA framework.

An Effective Entropy-Assisted Mind-Wandering Detection System Using EEG Signals of MM-SART Database.

Mind-wandering (MW), which is usually defined as a lapse of attention has negative effects on our daily life. Therefore, detecting when MW occurs can prevent us from those negative outcomes resulting from MW. In this work, we first collected a multi-modal Sustained Attention to Response Task (MM-SART) database for MW detection. Eighty-two participants' data were collected in our dataset. For each participant, we collected measures of 32-channels electroencephalogram (EEG) signals, photoplethysmography (PPG) signals, galvanic skin response (GSR) signals, eye tracker signals, and several questionnaires for detailed analyses. Then, we propose an effective MW detection system based on the collected EEG signals. To explore the non-linear characteristics of the EEG signals, we utilize entropy-based features. The experimental results show that we can reach 0.712 AUC score by using the random forest (RF) classifier with the leave-one-subject-out cross-validation. Moreover, to lower the overall computational complexity of the MW detection system, we propose correlation importance feature elimination (CIFE) along with AUC-based channel selection. By using two most significant EEG channels, we can reduce the training time of the classifier by 44.16%. By applying CIFE on the feature set, we can further improve the AUC score to 0.725 but with only 14.6% of the selection time compared with the recursive feature elimination (RFE). Finally, we can apply the current work to educational scenarios nowadays, especially in remote learning systems.

When Eyes Wander Around: Mind-Wandering as Revealed by Eye Movement Analysis with Hidden Markov Models.

Mind-wandering has been shown to largely influence our learning efficiency, especially in the digital and distracting era nowadays. Detecting mind-wandering thus becomes imperative in educational scenarios. Here, we used a wearable eye-tracker to record eye movements during the sustained attention to response task. Eye movement analysis with hidden Markov models (EMHMM), which takes both spatial and temporal eye-movement information into account, was used to examine if participants' eye movement patterns can differentiate between the states of focused attention and mind-wandering. Two representative eye movement patterns were discovered through clustering using EMHMM: centralized and distributed patterns. Results showed that participants with the centralized pattern had better performance on detecting targets and rated themselves as more focused than those with the distributed pattern. This study indicates that distinct eye movement patterns are associated with different attentional states (focused attention vs. mind-wandering) and demonstrates a novel approach in using EMHMM to study attention. Moreover, this study provides a potential approach to capture the mind-wandering state in the classroom without interrupting the ongoing learning behavior.

Coherence between Decomposed Components of Wrist and Finger PPG Signals by Imputing Missing Features and Resolving Ambiguous Features.

BACKGROUND: Feature extraction from photoplethysmography (PPG) signals is an essential step to analyze vascular and hemodynamic information. Different morphologies of PPG waveforms from different measurement sites appear. Various phenomena of missing or ambiguous features exist, which limit subsequent signal processing. METHODS: The reasons that cause missing or ambiguous features of finger and wrist PPG pulses are analyzed based on the concept of component waves from pulse decomposition. Then, a systematic approach for missing-feature imputation and ambiguous-feature resolution is proposed. RESULTS: From the experimental results, with the imputation and ambiguity resolution technique, features from 35,036 (98.7%) of 35,502 finger PPG cycles and 36307 (99.1%) of 36,652 wrist PPG cycles can be successfully identified. The extracted features became more stable and the standard deviations of their distributions were reduced. Furthermore, significant correlations up to 0.92 were shown between the finger and wrist PPG waveforms regarding the positions and widths of the third to fifth component waves. CONCLUSION: The proposed missing-feature imputation and ambiguous-feature resolution solve the problems encountered during PPG feature extraction and expand the feature availability for further processing. More intrinsic properties of finger and wrist PPG are revealed. The coherence between the finger and wrist PPG waveforms enhances the applicability of the wrist PPG.

Compressed-Domain ECG-Based Biometric User Identification Using Compressive Analysis.

Nowadays, user identification plays a more and more important role for authorized machine access and remote personal data usage. For reasons of privacy and convenience, biometrics-based user identification, such as iris, fingerprint, and face ID, has become mainstream methods in our daily lives. However, most of the biometric methods can be easily imitated or artificially cracked. New types of biometrics, such as electrocardiography (ECG), are based on physiological signals rather than traditional biological traits. Recently, compressive sensing (CS) technology that combines both sampling and compression has been widely applied to reduce the power of data acquisition and transmission. However, prior CS-based frameworks suffer from high reconstruction overhead and cannot directly align compressed ECG signals. In this paper, in order to solve the above two problems, we propose a compressed alignment-aided compressive analysis (CA-CA) algorithm for ECG-based biometric user identification. With CA-CA, it can avoid reconstruction and extract information directly from CS-based compressed ECG signals to reduce overall complexity and power. Besides, CA-CA can also align the compressed ECG signals in the eigenspace-domain, which can further enhance the precision of identifications and reduce the total training time. The experimental result shows that our proposed algorithm has a 94.16% accuracy based on a public database of 22 people.

Low-Complexity Privacy-Preserving Compressive Analysis Using Subspace-Based Dictionary for ECG Telemonitoring System.

Compressive sensing (CS) is attractive in long-term electrocardiography (ECG) telemonitoring to extend life-time for resource-constrained wireless wearable sensors. However, the availability of transmitted personal information has posed great concerns for potential privacy leakage. Moreover, the traditional CS-based security frameworks focus on secured signal recovery instead of privacy-preserving data analytics; hence, they provide only computational secrecy and have impractically high complexities for decryption. In this paper, to protect privacy from an information-theoretic perspective while delivering the classification capability, we propose a low-complexity framework of Privacy-Preserving Compressive Analysis (PPCA) based on subspace-based representation. The subspace-based dictionary is used for both encrypting and decoding the CS measurements online, and it is built by dividing signal space into discriminative and complementary subspace offline. The encrypted signal is unreconstructable even if the eavesdropper cracks the measurement matrix and the dictionary. PPCA is implemented in ECG-based atrial fibrillation detection. It can reduce the mutual information by 1.98 bits via encrypting measurements with signal-dependent noise at 1 dB, while the classification accuracy remains 96.05% with the decoding matrix. Furthermore, by decoding via matrix-vector product, rather than sparse coding, this computational complexity of PPCA is 341 times fewer compared with the traditional CS-based security.

Identification of Atrial Fibrillation by Quantitative Analyses of Fingertip Photoplethysmogram.

Atrial fibrillation (AF) detection is crucial for stroke prevention. We investigated the potential of quantitative analyses of photoplethysmogram (PPG) waveforms to identify AF. Continuous electrocardiogram (EKG) and fingertip PPG were recorded simultaneously in acute stroke patients (n = 666) admitted to an intensive care unit. Each EKG was visually labeled as AF (n = 150, 22.5%) or non-AF. Linear and nonlinear features from the pulse interval (PIN) and peak amplitude (AMP) of PPG waveforms were extracted from the first 1, 2, and 10 min of data. Logistic regression analysis revealed six independent PPG features feasibly identifying AF rhythm, including three PIN-related (mean, mean of standard deviation, and sample entropy), and three AMP-related features (mean of the root mean square of the successive differences, sample entropy, and turning point ratio) (all p < 0.01). The performance of the PPG analytic program comprising all 6 features that were extracted from the 2-min data was better than that from the 1-min data (area under the receiver operating characteristic curve was 0.972 (95% confidence interval 0.951-0.989) vs. 0.949 (0.929-0.970), p < 0.001 and was comparable to that from the 10-min data [0.973 (0.953-0.993)] for AF identification. In summary, our study established the optimal PPG analytic program in reliably identifying AF rhythm.

Complexity of heart rate variability predicts outcome in intensive care unit admitted patients with acute stroke .

BACKGROUND: Heart rate variability (HRV) has been proposed as a predictor of acute stroke outcome. This study aimed to evaluate the predictive value of a novel non-linear method for analysis of HRV, multiscale entropy (MSE) and outcome of patients with acute stroke who had been admitted to the intensive care unit (ICU). METHODS: The MSE of HRV was analysed from 1 h continuous ECG signals in ICU-admitted patients with acute stroke and controls. The complexity index was defined as the area under the MSE curve (scale 1-20). A favourable outcome was defined as modified Rankin scale 0-2 at 3 months after stroke. RESULTS: The trends of MSE curves in patients with atrial fibrillation (AF) (n=77) were apparently different from those in patients with non-AF stroke (n=150) and controls (n=60). In addition, the values of complexity index were significantly lower in the patients with non-AF stroke than in the controls (25.8±.3 vs. 32.3±4.3, p<0.001). After adjustment for clinical variables, patients without AF who had a favourable outcome were significantly related to higher complexity index values (OR=1.15, 95% CI 1.07 to 1.25, p<0.001). Importantly, the area under the receiver operating characteristic curve for predicting a favourable outcome of patients with non-AF stroke from clinical parameters was 0.858 (95% CI 0.797 to 0.919) and significantly improved to 0.903 (95% CI 0.853 to 0.954) after adding on the parameter of complexity index values (p=0.020). CONCLUSIONS: In ICU-admitted patients with acute stroke, early assessment of the complexity of HRV by MSE can help in predicting outcomes in patients without AF.

Effect of mannitol on cerebrovascular pressure reactivity in patients with intracranial hypertension.

BACKGROUND/PURPOSE: Mannitol is commonly used in patients with increased intracranial pressure (ICP), but its effect on cerebrovascular pressure reactivity (CVPR) is uncertain. We analyzed the changes of pressure reactivity index (PRx) during the course of mannitol treatment. METHODS: Twenty-one patients who received mannitol treatment for increased ICP were recruited prospectively. Continuous waveforms of arterial blood pressure (ABP) and ICP were collected simultaneously for 60 minutes (10 minutes at baseline and 50 minutes since mannitol administration) during 37 events of mannitol treatment. The correlation coefficients between the mean ABP and ICP were averaged every 10 minutes and labeled as the PRx. The linear correlation of six time points of PRx in each event was calculated to represent the trend of CVPR changes. The negative slope of correlation was defined as improvement in CVPR under mannitol treatment and vice versa. RESULTS: At baseline, the average of ICP was 26.0 ± 9.1 mmHg and the values of PRx were significantly correlated with ICP (p = 0.0044, r = 0.46). After mannitol administration, the average of ICP decreased significantly to 21.2 ± 11.1 mmHg (p = 0.036), and CVPR improved in 59.4 % of all events. Further analysis showed that low baseline cerebral perfusion pressure was the only hemodynamic parameter significant association with the improvement of CVPR after mannitol treatment (p = 0.039). CONCLUSION: Despite lowering ICP, mannitol may have diverse effects on CVPR in patients with intracranial hypertension. Our study suggests that mannitol infusion may have a beneficial effect on CVPR, particularly in those with a low cerebral perfusion pressure at baseline.