[Electrocardiogram signal classification algorithm of nested long short-term memory network based on focal loss function].

Electrocardiogram (ECG) can visually reflect the physiological electrical activity of human heart, which is important in the field of arrhythmia detection and classification. To address the negative effect of label imbalance in ECG data on arrhythmia classification, this paper proposes a nested long short-term memory network (NLSTM) model for unbalanced ECG signal classification. The NLSTM is built to learn and memorize the temporal characteristics in complex signals, and the focal loss function is used to reduce the weights of easily identifiable samples. Then the residual attention mechanism is used to modify the assigned weights according to the importance of sample characteristic to solve the sample imbalance problem. Then the synthetic minority over-sampling technique is used to perform a simple manual oversampling process on the Massachusetts institute of technology and Beth Israel hospital arrhythmia (MIT-BIH-AR) database to further increase the classification accuracy of the model. Finally, the MIT-BIH arrhythmia database is applied to experimentally verify the above algorithms. The experimental results show that the proposed method can effectively solve the issues of imbalanced samples and unremarkable features in ECG signals, and the overall accuracy of the model reaches 98.34%. It also significantly improves the recognition and classification of minority samples and has provided a new feasible method for ECG-assisted diagnosis, which has practical application significance.

Resting-State MEG Source Space Network Metrics Associated with the Duration of Temporal Lobe Epilepsy.

To evaluate the relationship between the network metrics of 68 brain regions and duration of temporal lobe epilepsy (TLE). Magnetoencephalography (MEG) data from 53 patients with TLE (28 left TLE, 25 right TLE) were recorded between seizures at resting state and analyzed in six frequency bands: delta (0.1-4 Hz), theta (4-8 Hz), lower alpha (8-10 Hz), upper alpha (10-13 Hz), beta (13-30 Hz), and lower gamma (30-48 Hz). Three local network metrics, betweenness centrality, nodal degree, and nodal efficiency, were chosen to analyze the functional brain network. In Left, Right, and All (Left + Right) TLE groups, different metrics provide significant positive or negative correlations with the duration of TLE, in different frequency bands, and in different brain regions. In the Left TLE group, significant correlation between TLE duration and metric exists in the delta, beta, or lower gamma band, with network betweenness centrality, nodal degree, or nodal efficiency, in left caudal middle frontal, left middle temporal, or left supramarginal. In the Right TLE group, significant correlation exists in lower gamma or delta band, with nodal degree, or nodal efficiency, in left precuneus or right temporal pole. In the All TLE group, the significant correlation exists in delta, theta, beta, or lower gamma band, with nodal degree, or betweenness centrality, in either left or right hemisphere. Network metrics for some specific brain regions changed in patients with TLE as the duration of their TLE increased. Further researching these changes may be important for studying the pathogenesis, presurgical evaluation, and clinical treatment of long-term TLE.

[Research on high-efficiency electrocardiogram automatic classification based on autoregressive moving average model fitting].

The automatic detection of arrhythmia is of great significance for the early prevention and diagnosis of cardiovascular diseases. Traditional arrhythmia diagnosis is limited by expert knowledge and complex algorithms, and lacks multi-dimensional feature representation capabilities, which is not suitable for wearable electrocardiogram (ECG) monitoring equipment. This study proposed a feature extraction method based on autoregressive moving average (ARMA) model fitting. Different types of heartbeats were used as model inputs, and the characteristic of fast and smooth signal was used to select the appropriate order for the arrhythmia signal to perform coefficient fitting, and complete the ECG feature extraction. The feature vectors were input to the support vector machine (SVM) classifier and K-nearest neighbor classifier (KNN) for automatic ECG classification. MIT-BIH arrhythmia database and MIT-BIH atrial fibrillation database were used to verify in the experiment. The experimental results showed that the feature engineering composed of the fitting coefficients of the ARMA model combined with the SVM classifier obtained a recall rate of 98.2% and a precision rate of 98.4%, and the F 1 index was 98.3%. The algorithm has high performance, meets the needs of clinical diagnosis, and has low algorithm complexity. It can use low-power embedded processors for real-time calculations, and it's suitable for real-time warning of wearable ECG monitoring equipment.

Data on flexibility and thermal stability of Polypyrrole-based ternary nanocomposite films.

Carbon materials, such as multi-walled carbon nanotube (MWCNT), single-walled carbon nanotubes (SWCNT), and graphene sheets (GNS), filling into polymer substrates can effectively improve performance of composite materials [1], [2]. The data presented here in this article illustrates the different impacts of GNS and MWCNT on the mechanical properties of polypyrrole (PPy)-based composites systems. PPy/GNS and PPy/MWCNT binary composites were added into poly(3,4-ethylenedioxythiophene): poly (styrene sulfonate) (PEDOT: PSS) matrix. Changing the ratio of PPy/GNS and PPy/MWCNT to PEDOT: PSS, a series of PEDOT: PSS-PPy/GNS (abbreviated as PGNS) and PEDOT: PSS-PPy/MWCNT (abbreviated as PMWCNT) ternary composites films were obtained. The synthesis process of PGNS and PMWCNT films are based on Wang et al [3].

Novel Hybrid p- and n-Type Organic Thermoelectric Materials Based on Mussel-Inspired Polydopamine.

Mussel-inspired polydopamine (PDA) can serve as building blocks and interfaces in designing functional materials. Here, the use of PDA as an interlayer between polyaniline (PANi) and multidimensional carbon materials, such as graphene quantum dots (GQD), multiwalled carbon nanotubes (MWCNT), and graphene nanosheets (GNS), to improve the thermoelectric performance of p-type polymer-based materials has been reported. The introduction of PDA promotes the carrier mobility of GQD/PDA/PANi, CNT/PDA/PANi, and GNS/PDA/PANi ternary composites due to the superior adhesive property of PDA. An optimal conductivity of 4.98 × 104 S m-1 and a power factor of 92.17 µW m-1 K-2 at 363 K are achieved in GNS/PDA/PANi, which are much higher than the values of GNS/PDA and GNS/PANi. More surprisingly, despite the fact that GQD/PDA, CNT/PDA, and GNS/PDA binary composites show p-type properties, the pyrolysis treatment of GQD/PDA, CNT/PDA, and GNS/PDA at 800 °C results in a gain in both the electrical conductivity and negative Seebeck coefficient of c-GQD/PDA, c-CNT/PDA, and c-GNS/PDA. The c-CNT/PDA composites possess the highest Seebeck value of -30.2 µV K-1 and a maximum power factor value of 35.57 µW m-1 K-2. Finally, a flexible thermoelectric generator with 24 thermoelectric units composed of GNS/PDA/PANi and c-CNT/PDA is demonstrated, which gives an output voltage of 52.8 mV at a temperature difference of 60 °C.

A Look Into the Power of fNIRS Signals by Using the Welch Power Spectral Estimate for Deception Detection.

Neuroimaging technologies have improved our understanding of deception and also exhibit their potential in revealing the origins of its neural mechanism. In this study, a quantitative power analysis method that uses the Welch power spectrum estimation of functional near-infrared spectroscopy (fNIRS) signals was proposed to examine the brain activation difference between the spontaneous deceptive behavior and controlled behavior. The power value produced by the model was applied to quantify the activity energy of brain regions, which can serve as a neuromarker for deception detection. Interestingly, the power analysis results generated from the Welch spectrum estimation method demonstrated that the spontaneous deceptive behavior elicited significantly higher power than that from the controlled behavior in the prefrontal cortex. Meanwhile, the power findings also showed significant difference between the spontaneous deceptive behavior and controlled behavior, indicating that the reward system was only involved in the deception. The proposed power analysis method for processing fNIRS data provides us an additional insight to understand the cognitive mechanism of deception.

Influence of EEG References on N170 Component in Human Facial Recognition.

The choice of the reference electrode scheme is an important step in event-related potential (ERP) analysis. In order to explore the optimal electroencephalogram reference electrode scheme for the ERP signal related to facial recognition, we investigated the influence of average reference (AR), mean mastoid reference (MM), and Reference Electrode Standardization Technique (REST) on the N170 component via statistical analysis, statistical parametric scalp mappings (SPSM) and source analysis. The statistical results showed that the choice of reference electrode scheme has little effect on N170 latency (p > 0.05), but has an significant impact on N170 amplitude (p < 0.05). ANOVA results show that, for the three references scheme, there was statistically significant difference between N170 latency and amplitude induced by the unfamiliar face and that induced by the scrambled face (p < 0.05). Specifically, the SPSM results show an anterior and a temporo-occipital distribution for AR and REST, whereas just anterior distribution shown for MM. However, the circumstantial evidence provided by source analysis is more consistent with SPSM of AR and REST, compared with that of MM. These results indicate that the experimental results under the AR and REST references are more objective and appropriate. Thus, it is more appropriate to use AR and REST reference scheme settings in future facial recognition experiments.