No full-text available
To read the full-text of this research,
you can request a copy directly from the author.
Abstract
Artificial intelligence (AI) is increasingly being used in clinical anesthesia, and researchers are using algorithms to dig information from patients’ perioperative data, process and analyze them from multi-dimensions, after which predictive models are built to dynamically predict perioperative adverse events.
ResearchGate has not been able to resolve any citations for this publication.
Emotion recognition can be achieved by speech recognition, the judgment of limb movements, analysis of Electrooculogram (EOG) or capturing of facial expressions. However, those types of emotion recognition methods cannot detect human emotion well, because humankind can use fake body movement and words to hide real emotions. In this paper, we proposed an EEG-based emotion classification method based on Bidirectional Long Short-Term Memory Network (BiLSTM). Electroencephalogram (EEG) signal can detect human emotion correctly because human represent their real emotions in their mind and cannot hide emotions there. Meanwhile, EEG is a time sequence signal which needs a model which can deal with this type of data. Therefore, we chose Long Short-term Memory Network to process the EEG signal. In particular, we used an improvement version of LSTM model BiLSTM to manage the signals. BiLSTM can processes input data from front to back and back to front. Meanwhile, BiLSTM can store important information and forget unnecessary information; therefore, this process increases the accuracy of the model. Our method classifies four discrete classifications (happy, sad, fear, and neutral) for emotion classification, which achieves competitive performance compared with other conventional emotion classification methods. The final experimental results show that we can achieve an accuracy of 84.21% for four emotional states classification by using our method.
The identification of sleep stages is essential in the diagnostics of sleep disorders, among which obstructive sleep apnea (OSA) is one of the most prevalent. However, manual scoring of sleep stages is time-consuming, subjective, and costly. To overcome this shortcoming, we aimed to develop an accurate deep learning approach for automatic classification of sleep stages and to study the effect of OSA severity on the classification accuracy. Overnight polysomnographic recordings from a public dataset of healthy individuals (Sleep-EDF, n=153) and from a clinical dataset (n=891) of patients with suspected OSA were used to develop a combined convolutional and long short-term memory neural network. On the public dataset, the model achieved sleep staging accuracy of 83.7% (κ=0.77) with a single frontal EEG channel and 83.9% (κ=0.78) when supplemented with EOG. For the clinical dataset, the model achieved accuracies of 82.9% (κ=0.77) and 83.8% (κ=0.78) with a single EEG channel and two channels (EEG+EOG), respectively. The sleep staging accuracy decreased with increasing OSA severity. The single-channel accuracy ranged from 84.5% (κ=0.79) for individuals without OSA diagnosis to 76.5% (κ=0.68) for severe OSA patients. In conclusion, deep learning enables automatic sleep staging for suspected OSA patients with high accuracy and expectedly, the accuracy lowered with increasing OSA severity. Furthermore, the accuracies achieved in the public dataset were superior to previously published state-of-the-art methods. Adding an EOG channel did not significantly increase the accuracy. The automatic, single-channel-based sleep staging could enable easy, accurate, and cost-efficient integration of EEG recording into diagnostic ambulatory recordings.
One of the most challenging predictive data analysis efforts is accurate prediction of depth of anesthesia (DOA) indicators which has attracted a growing attention since it provides patients a safe surgical environment in case of secondary damage caused by intraoperative awareness or brain injury. However, many researchers put heavily handcraft feature extraction or carefully tailored feature engineering to each patient to achieve very high sensitivity and low false prediction rate for a particular dataset. This limits the benefit of the proposed approaches if a different dataset is used. Recently, representations learned using deep convolutional neural network (CNN) for object recognition are becoming widely used model of the processing hierarchy in the human visual system. The correspondence between models and brain signals that holds the acquired activity at high temporal resolution has been explored less exhaustively. In this paper, deep learning CNN with a range of different architectures, is designed for identifying related activities from raw electroencephalography (EEG). Specifically, an improved short-time Fourier transform (STFT) is used to stand for the time-frequency information after extracting the spectral images of the original EEG as input to CNN. Then CNN models are designed and trained to predict the DOA levels from EEG spectrum without handcrafted features, which presents an intuitive mapping process with high efficiency and reliability. As a result, the best trained CNN model achieved an accuracy of 93.50%, interpreted as CNN’s deep learning to approximate the DOA by senior anesthesiologists, which highlights the potential of deep CNN combined with advanced visualization techniques for EEG-based brain mapping.
Polycyclic aromatic compounds (PACs) are known due to their mutagenic activity. Among them, 2-nitrobenzanthrone (2-NBA) and 3-nitrobenzanthrone (3-NBA) are considered as two of the most potent mutagens found in atmospheric particles. In the present study 2-NBA, 3-NBA and selected PAHs and Nitro-PAHs were determined in fine particle samples (PM 2.5) collected in a bus station and an outdoor site. The fuel used by buses was a diesel-biodiesel (96:4) blend and light-duty vehicles run with any ethanol-to-gasoline proportion. The concentrations of 2-NBA and 3-NBA were, on average, under 14.8 µg g⁻¹ and 4.39 µg g⁻¹, respectively. In order to access the main sources and formation routes of these compounds, we performed ternary correlations and multivariate statistical analyses. The main sources for the studied compounds in the bus station were diesel/biodiesel exhaust followed by floor resuspension. In the coastal site, vehicular emission, photochemical formation and wood combustion were the main sources for 2-NBA and 3-NBA as well as the other PACs. Incremental lifetime cancer risk (ILCR) were calculated for both places, which presented low values, showing low cancer risk incidence although the ILCR values for the bus station were around 2.5 times higher than the ILCR from the coastal site.
Guidelines are presented for safe practice in the use of intravenous drug infusions for general anaesthesia. When maintenance of general anaesthesia is by intravenous infusion, this is referred to as total intravenous anaesthesia. Although total intravenous anaesthesia has advantages for some patients, the commonest technique used for maintenance of anaesthesia in the UK and Ireland remains the administration of an inhaled volatile anaesthetic. However, the use of an inhalational technique is sometimes not possible, and in some situations, inhalational anaesthesia is contraindicated. Therefore, all anaesthetists should be able to deliver total intravenous anaesthesia competently and safely. For the purposes of simplicity, these guidelines will use the term total intravenous anaesthesia but also encompass techniques involving a combination of intravenous infusion and inhalational anaesthesia. This document is intended as a guideline for safe practice when total intravenous anaesthesia is being used, and not as a review of the pros and cons of total intravenous anaesthesia vs. inhalational anaesthesia in situations where both techniques are possible.
Brain Computer-Interfacing is a methodology that provides a way for communication with the outside environment using the brain thoughts. The success of this methodology depends on the selection of methods to process the brain signals in each phase. This paper aimed at addressing the various methodologies required to be adapted in each phase of brain signal processing. Prior to this survey, previous surveys have been listed various methods, some experimental results and compared them. This paper shows clear and easy interpretation of each method and their advantages and disadvantages including the signal acquisition, signal enhancement, feature extraction and, signal classification.
This paper describes the classification of facial expressions using EEG data. The entire procedure aims at controlling an electric wheelchair with a Brain Computer Interface (BCI) headset. The goal is to help the people who are suffering from locked-in syndrome to move or to pass the necessary signals. The headset consists of the electroencephalogram (EEG) cap comprising 16 electrodes attached to the amplifier out of which 14 electrodes are used for data acquisition while two are used as reference and ground. The EEG cap is placed on the head of the subject and various expressions (blink, eyebrows raise, smile, etc.) are performed on the subject. Muscle activities due to facial expressions can be observed from the recorded EEG signals. Expressions are classified and necessary signals are generated. The features are extracted using the wavelet packet transform processing method and classified primarily using Support Vector Machine (SVM).
Complexity science has provided new perspectives and opportunities for understanding a variety of complex natural or social phenomena, including brain dysfunctions like epilepsy. By delving into the complexity in electrophysiological signals and neuroimaging, new insights have emerged. These discoveries have revealed that complexity is a fundamental aspect of physiological processes. The inherent nonlinearity and non-stationarity of physiological processes limits the methods based on simpler underlying assumptions to point out the pathway to a more comprehensive understanding of their behavior and relation with certain diseases. The perspective of complexity may benefit both the research and clinical practice through providing novel data analytics tools devoted for the understanding of and the intervention about epilepsies. This review aims to provide a sketchy overview of the methods derived from different disciplines lucubrating to the complexity of bio-signals in the field of epilepsy monitoring. Although the complexity of bio-signals is still not fully understood, bundles of new insights have been already obtained. Despite the promising results about epileptic seizure detection and prediction through offline analysis, we are still lacking robust, tried-and-true real-time applications. Multidisciplinary collaborations and more high-quality data accessible to the whole community are needed for reproducible research and the development of such applications.
Estimating the depth of anaesthesia (DoA) in operations has always been a challenging issue due to the underlying complexity of the brain mechanisms. Electroencephalogram (EEG) signals are undoubtedly the most widely used signals for measuring DoA. In this paper, a novel EEG-based index is proposed to evaluate DoA for 24 patients receiving general anaesthesia with different levels of unconsciousness. Sample Entropy (SampEn) algorithm was utilised in order to acquire the chaotic features of the signals. After calculating the SampEn from the EEG signals, Random Forest was utilised for developing learning regression models with Bispectral index (BIS) as the target. Correlation coefficient, mean absolute error, and area under the curve (AUC) were used to verify the perioperative performance of the proposed method. Validation comparisons with typical nonstationary signal analysis methods (i.e., recurrence analysis and permutation entropy) and regression methods (i.e., neural network and support vector machine) were conducted. To further verify the accuracy and validity of the proposed methodology, the data is divided into four unconsciousness-level groups on the basis of BIS levels. Subsequently, analysis of variance (ANOVA) was applied to the corresponding index (i.e., regression output). Results indicate that the correlation coefficient improved to 0.72 ± 0.09 after filtering and to 0.90 ± 0.05 after regression from the initial values of 0.51 ± 0.17. Similarly, the final mean absolute error dramatically declined to 5.22 ± 2.12. In addition, the ultimate AUC increased to 0.98 ± 0.02, and the ANOVA analysis indicates that each of the four groups of different anaesthetic levels demonstrated significant difference from the nearest levels. Furthermore, the Random Forest output was extensively linear in relation to BIS, thus with better DoA prediction accuracy. In conclusion, the proposed method provides a concrete basis for monitoring patients’ anaesthetic level during surgeries.
Seizure prediction has attracted growing attention as one of the most challenging predictive data analysis efforts to improve the life of patients with drug-resistant epilepsy and tonic seizures. Many outstanding studies have reported great results in providing sensible indirect (warning systems) or direct (interactive neural stimulation) control over refractory seizures, some of which achieved high performance. However, to achieve high sensitivity and a low false prediction rate, many of these studies relied on handcraft feature extraction and/or tailored feature extraction, which is performed for each patient independently. This approach, however, is not generalizable, and requires significant modifications for each new patient within a new dataset. In this article, we apply convolutional neural networks to different intracranial and scalp electroencephalogram (EEG) datasets and propose a generalized retrospective and patient-specific seizure prediction method. We use the short-time Fourier transform on 30-s EEG windows to extract information in both the frequency domain and the time domain. The algorithm automatically generates optimized features for each patient to best classify preictal and interictal segments. The method can be applied to any other patient from any dataset without the need for manual feature extraction. The proposed approach achieves sensitivity of 81.4%, 81.2%, and 75% and a false prediction rate of 0.06/h, 0.16/h, and 0.21/h on the Freiburg Hospital intracranial EEG dataset, the Boston Children's Hospital-MIT scalp EEG dataset, and the American Epilepsy Society Seizure Prediction Challenge dataset, respectively. Our prediction method is also statistically better than an unspecific random predictor for most of the patients in all three datasets.
Recent advancements in human–computer interaction research has led to the possibility of emotional communication via brain–computer interface systems for patients with neuropsychiatric disorders or disabilities. In this study, we efficiently recognize emotional states by analyzing the features of electroencephalography (EEG) signals, which are generated from EEG sensors that non-invasively measure the electrical activity of neurons inside the human brain, and select the optimal combination of these features for recognition. In this study, the scalp EEG data of 21 healthy subjects (12–14 years old) were recorded using a 14-channel EEG machine while the subjects watched images with four types of emotional stimuli (happy, calm, sad, or scared). After preprocessing, the Hjorth parameters (activity, mobility, and complexity) were used to measure the signal activity of the time series data. We selected the optimal EEG features using a balanced one-way ANOVA after calculating the Hjorth parameters for different frequency ranges. Features selected by this statistical method outperformed univariate and multivariate features. The optimal features were further processed for emotion classification using support vector machine (SVM), k-nearest neighbor (KNN), linear discriminant analysis (LDA), Naive Bayes, Random Forest, deep-learning, and four ensembles methods (bagging, boosting, stacking, and voting). The results show that the proposed method substantially improves the emotion recognition rate with respect to the commonly used spectral power band method.
Objective:
Considering the importance and the near future development of noninvasive Brain-Machine Interface (BMI) systems, this paper presents a comprehensive theoretical-experimental survey on the classification and evolutionary methods for BMI-based systems in which EEG signals are used.
Approach:
The paper is divided into two main parts. In the first part a wide range of different types of the base and combinatorial classifiers including boosting and bagging classifiers and also evolutionary algorithms are reviewed and investigated. In the second part, these classifiers and evolutionary algorithms are assessed and compared based on two types of relatively widely used BMI systems, that is, Sensory Motor Rhythm-BMI (SMR-BMI) and Event Related Potentials-BMI (ERPs-BMI). Moreover, in the second part, some of the improved evolutionary algorithms as well as bi-objective algorithms are experimentally assessed and compared.
Main results:
In this study two databases are used, and cross-validation accuracy (CVA) and stability to data volume (SDV) are considered as the evaluation criteria for the classifiers. According to the experimental results on both databases, regarding the base classifiers, LDA (Linear Discriminant Analysis) and SVM (Support Vector Machines) with respect to CVA evaluation metric, and NB (Naive Bayes) with respect to SDV demonstrated the best performances. Among the combinatorial classifiers, four classifiers Bagg-DT (Bagging Decision Tree), LogitBoost, and GentleBoost with respect to CVA, and Bagging-LR (Bagging Logistic Regression) and AdaBoost (Adaptive Boosting) with respect to SDV had the best performances. Finally, regarding the evolutionary algorithms, single-objective IWO (Invasive Weed Optimization) and bi-objective NSIWO (Nondominated Sorting IWO) algorithms demonstrated the best performances.
Significance:
We present a general survey on the base and the combinatorial classification methods for EEG signals (sensory motor rhythm and event related potentials) as well as their optimization methods through the evolutionary algorithms. In addition, experimental and statistical significance tests are carried out to study the applicability and effectiveness of the reviewed methods.
Background. Postoperative residual curarization (PORC) after surgery is common and its detection has a high error rate. Artificial neural networks are being used increasingly to examine complex data. We hypothesized that a neural network would enhance prediction of PORC.
Methods. In 40 previously reported patients, neuromuscular function, neuromuscular block/antagonist usage and time intervals were recorded throughout anaesthesia until tracheal extubation by an observer uninvolved in patient care. PORC was defined as significant ‘fade’ (train of four <0.7) at extubation. Neuromuscular function was classified as PORC (value=1) or no PORC (value=0). A back‐propagation neural network was trained to assign similar values (0, 1) for prediction of PORC, by examining the impact of (i) the degree of spontaneous recovery at reversal, and (ii) the time since pharmacological reversal, using the jackknife method. Successful prediction was defined as attainment of a predicted value within 0.2 of the target value.
Results. Twenty‐six patients (65%) had PORC at tracheal extubation. Clinical detection of PORC had a sensitivity of 0 and specificity of 1, with an indeterminate positive predictive value and a negative predictive value of 0.35. Using the artificial neural network, one patient with residual block and one with adequate neuromuscular function were incorrectly classified during the test phase, with no indeterminate predictions, giving an artificial neural network sensitivity of 0.96 (χ²=44, P<0.001) and specificity of 0.92 (P=1), with a positive predictive value of 0.96 and a negative predictive value of 0.93 (χ²=12, P<0.001).
Conclusions. Neural network‐based prediction, using readily available clinical measurements, is significantly better than human judgement in predicting recovery of neuromuscular function.
Br J Anaesth 2003; 90: 48–52
With a massive influx of multimodality data, the role of data analytics in health informatics has grown rapidly in the last decade. This has also prompted increasing interests in the generation of analytical, data driven models based on machine learning in health informatics. Deep learning, a technique with its foundation in artificial neural networks, is emerging in recent years as a powerful tool for machine learning, promising to reshape the future of artificial intelligence. Rapid improvements in computational power, fast data storage and parallelization have also contributed to the rapid uptake of the technology in addition to its predictive power and ability to generate automatically optimized high-level features and semantic interpretation from the input data. This article presents a comprehensive up-to-date review of research employing deep learning in health informatics, providing a critical analysis of the relative merit and potential pitfalls of the technique as well as its future outlook. The paper mainly focuses on key applications of deep learning in the fields of translational bioinformatics, medical imaging, pervasive sensing, medical informatics and public health.
Determining depth of anesthesia is a challenging problem in the context of biomedical signal processing. Various methods have been suggested to determine a quantitative index as depth of anesthesia, but most of these methods suffer from high sensitivity during the surgery. A novel method based on energy scattering of samples in the wavelet domain is suggested to represent the basic content of electroencephalogram (EEG) signal. In this method, first EEG signal is decomposed into different sub-bands, then samples are squared and energy of samples sequence is constructed through each scale and time, which is normalized and finally entropy of the resulted sequences is suggested as a reliable index. Empirical Results showed that applying the proposed method to the EEG signals can classify the awake, moderate and deep anesthesia states similar to BIS.
This study evaluated the depth of anesthesia (DoA) index using artificial neural networks (ANN) which is performed as the modeling technique. Totally 63-patient data is addressed, for both modeling and testing of 17 and 46 patients, respectively. The empirical mode decomposition (EMD) is utilized to purify between the electroencephalography (EEG) signal and the noise. The filtered EEG signal is subsequently extracted to achieve a sample entropy index by every 5-second signal. Then, it is combined with other mean values of vital signs, that is, electromyography (EMG), heart rate (HR), pulse, systolic blood pressure (SBP), diastolic blood pressure (DBP), and signal quality index (SQI) to evaluate the DoA index as the input. The 5 doctor scores are averaged to obtain an output index. The mean absolute error (MAE) is utilized as the performance evaluation. 10-fold cross-validation is performed in order to generalize the model. The ANN model is compared with the bispectral index (BIS). The results show that the ANN is able to produce lower MAE than BIS. For the correlation coefficient, ANN also has higher value than BIS tested on the 46-patient testing data. Sensitivity analysis and cross-validation method are applied in advance. The results state that EMG has the most effecting parameter, significantly.
Recently, several attempts have been made to find the depth of anesthesia (DOA) by analyzing the ongoing electroencephalogram (EEG) signals during surgical operations. Nevertheless, specialists still do not rely on these indexes because they cannot accurately track the transitions of anesthetic depth. This paper presents an effective EEG-based index that is fast to compute and acts very accurate in practice. To determine the proposed index, first EEG signals are denoised with an adaptive thresholding method. The wavelet transform is then applied to the clean EEG signals in order to decompose the signal into brain-match subspaces and the proposed feature extracted from each subspace to monitor the DOA. EEG signals of 8 subjects were recorded during the surgical operation. Experimental results exhibit the proposed features highly correlated with the BIS index (the most popular EEG-based index)through different anesthetic levels. Moreover, in some cases the introduced index outperformed the BIS and the clinical observation confirmed this superiority.
Chinese named entity recognition is a crucial initial step of information extraction in the field of agricultural diseases and pests. This step aims to identify named entities related to agricultural diseases and pests from unstructured texts but presents challenges. The available corpus in this domain is limited, and most existing named entity recognition methods only focus on the global context information but neglect potential local context features, which are also equally important for named entity recognition. To solve the above problems and tackle the named entity recognition task in this paper, an available corpus toward agricultural diseases and pests, namely AgCNER, which contains 11 categories and 34,952 samples, was established. Compared with the corpora in the same field, this corpus has additional categories and more sample sizes. Then, a novel Chinese named entity recognition model via joint multi-scale local context features and the self-attention mechanism was proposed. The original Bi-directional Long Short-Term Memory and Conditional Random Field model (BiLSTM-CRF) was improved by fusing the multi-scale local context features extracted by Convolutional Neural Network (CNN) with different kernel sizes. The self-attention mechanism was also used to break the limitation of BiLSTM-CRF in capturing long-distance dependencies and further improve the model performance. The performance of the proposed model was evaluated on three corpora, namely AgCNER, Resume, and MSRA, which achieved the optimal F1-values of 94.15%, 94.56%, and 90.55%, respectively. Experimental results in many aspects illustrated the effective performance of the proposed model in this paper.
We present an end-to-end deep learning model that can automatically detect epileptic seizures in multichannel electroencephalography (EEG) recordings. Our model combines a Convolutional Neural Network (CNN) and a Bidirectional Long Short-Term Memory (BLSTM) network to efficiently mine information from the EEG data using a small number of trainable parameters. Specifically, the CNN learns a latent encoding for each one second window of raw multichannel EEG data. In conjunction, the BLSTM learns the temporal evolution of seizure presentations given the CNN encodings. The combination of these architectures allows our model to capture both the short time scale EEG features indicative of seizure activity as well as the long term correlations in seizure presentations. Unlike most prior work in seizure detection, we mimic an in-patient monitoring setting through a leave-one-patient-out cross validation procedure, attaining an average seizure detection sensitivity of 0.91 across all patients. This strategy verifies that our model can generalize to new patients. We demonstrate that our CNN–BLSTM outperforms both conventional feature extraction methods and state-of-the-art deep learning approaches that rely on larger and more complex network architectures.
Objective:
To develop an effective and scalable individual-level patient cost prediction method by automatically learning hidden temporal patterns from multivariate time series data in patient insurance claims using a convolutional neural network (CNN) architecture.
Methods:
We used three years of medical and pharmacy claims data from 2013 to 2016 from a healthcare insurer, where data from the first two years were used to build the model to predict costs in the third year. The data consisted of the multivariate time series of cost, visit and medical features that were shaped as images of patients' health status (i.e., matrices with time windows on one dimension and the medical, visit and cost features on the other dimension). Patients' multivariate time series images were given to a CNN method with a proposed architecture. After hyper-parameter tuning, the proposed architecture consisted of three building blocks of convolution and pooling layers with an LReLU activation function and a customized kernel size at each layer for healthcare data. The proposed CNN learned temporal patterns became inputs to a fully connected layer. We benchmarked the proposed method against three other methods: 1) a spike temporal pattern detection method, as the most accurate method for healthcare cost prediction described to date in the literature; 2) a symbolic temporal pattern detection method, as the most common approach for leveraging healthcare temporal data; and 3) the most commonly used CNN architectures for image pattern detection (i.e., AlexNet, VGGNet and ResNet) (via transfer learning). Moreover, we assessed the contribution of each type of data (i.e., cost, visit and medical). Finally, we externally validated the proposed method against a separate cohort of patients. All prediction performances were measured in terms of mean absolute percentage error (MAPE).
Results:
The proposed CNN configuration outperformed the spike temporal pattern detection and symbolic temporal pattern detection methods with a MAPE of 1.67 versus 2.02 and 3.66, respectively (p<0.01). The proposed CNN outperformed ResNet, AlexNet and VGGNet with MAPEs of 4.59, 4.85 and 5.06, respectively (p<0.01). Removing medical, visit and cost features resulted in MAPEs of 1.98, 1.91 and 2.04, respectively (p<0.01).
Conclusions:
Feature learning through the proposed CNN configuration significantly improved individual-level healthcare cost prediction. The proposed CNN was able to outperform temporal pattern detection methods that look for a pre-defined set of pattern shapes, since it is capable of extracting a variable number of patterns with various shapes. Temporal patterns learned from medical, visit and cost data made significant contributions to the prediction performance. Hyper-parameter tuning showed that considering three-month data patterns has the highest prediction accuracy. Our results showed that patients' images extracted from multivariate time series data are different from regular images, and hence require unique designs of CNN architectures. The proposed method for converting multivariate time series data of patients into images and tuning them for convolutional learning could be applied in many other healthcare applications with multivariate time series data.
Background
The electrocardiogram (ECG) is one of the most commonly used diagnostic tools in medicine and healthcare. Deep learning methods have achieved promising results on predictive healthcare tasks using ECG signals.
Objective
This paper presents a systematic review of deep learning methods for ECG data from both modeling and application perspectives.
Methods
We extracted papers that applied deep learning (deep neural network) models to ECG data that were published between January 1st of 2010 and February 29th of 2020 from Google Scholar, PubMed, and the Digital Bibliography & Library Project. We then analyzed each article according to three factors: tasks, models, and data. Finally, we discuss open challenges and unsolved problems in this area.
Results
The total number of papers extracted was 191. Among these papers, 108 were published after 2019. Different deep learning architectures have been used in various ECG analytics tasks, such as disease detection/classification, annotation/localization, sleep staging, biometric human identification, and denoising.
Conclusion
The number of works on deep learning for ECG data has grown explosively in recent years. Such works have achieved accuracy comparable to that of traditional feature-based approaches and ensembles of multiple approaches can achieve even better results. Specifically, we found that a hybrid architecture of a convolutional neural network and recurrent neural network ensemble using expert features yields the best results. However, there are some new challenges and problems related to interpretability, scalability, and efficiency that must be addressed. Furthermore, it is also worth investigating new applications from the perspectives of datasets and methods.
Significance
This paper summarizes existing deep learning research using ECG data from multiple perspectives and highlights existing challenges and problems to identify potential future research directions.
There are a lot of linguistic knowledge and sentiment resources nowadays, but in the current research with deep learning framework, these kinds of unique sentiment information are not fully used in sentiment analysis tasks. Moreover, the sentiment analysis task can be seen as a sequence model, and the sequence model has a problem: the model will decode the input file sequences into a specific length vector. If the length of the vectors is set too short, the input text information will be lost, and finally the text will be misjudged. To solve these problems, we propose a bidirectional LSTM model with self-attention mechanism and multi-channel features (SAMF-BiLSTM). The method models the existing linguistic knowledge and sentiment resources in sentiment analysis tasks to form different feature channels, and uses self-attention mechanism to enhance the sentiment information. SAMF-BiLSTM model can fully exploit the relationship between target words and sentiment polarity words in a sentence, and does not rely on manually organized sentiment lexicon. In addition, we propose the SAMF-BiLSTM-D model based on SAMF-BiLSTM model for document-level text classification tasks. The method obtains the representation of all sentences in the document through SAMF-BiLSTM training, then integrates BiLSTM to learn the representation of all sentences, and further obtains the sentiment feature information of the entire document. Finally, we evaluate experiment results under five datasets. The results show that SAMF-BiLSTM and SAMF-BiLSTM-D are superior to other advanced methods in classification accuracy in most cases.
This comprehensive, evidence-based book is intended to serve as a reference for medical practitioners involved in the perioperative care of neurosurgical patients. Fundamental aspects of neuroanesthesiology and neurocritical care are thoroughly examined across 101 chapters, outlining key elements that are crucial to a care provider’s knowledge of the practice. These elements include specific diagnostic procedures and treatment options, concepts and applicable details of the available neurosurgical interventions and techniques, and mechanisms necessary to provide top of the line care for patients. Each chapter features definitive and distinct areas of this multi-specialty discipline, and is designed to guide the reader from problem to solution in situations that can arise in the clinical setting.
Essentials of Neurosurgical Anesthesia & Critical Care, 2nd edition is a problem-oriented approach textbook that will aid a wide variety of readers in handling day-to-day issues and developments relevant to the perioperative care of neurosurgical patients.
Multivariate time series prediction, with a profound impact on human social life, has been attracting growing interest in machine learning research. However, the task of time series forecasting is very challenging because it is affected by many complex factors. For example, in predicting traffic and solar power generation, weather can bring great trouble. In particular, for strictly periodic time series, if the periodic information can be extracted from the historical sequence data to the maximum, the accuracy of the prediction will be greatly improved. At present, for time series prediction tasks, the sequence models based on RNN have made great progress. However, the sequence models has difficulty in capturing global information, failing to well highlight the periodic characteristics of the time series. But the this problem can be solved by CNN models. So in this paper, we propose a model called Multiple CNNs to solve the problem of periodic multivariate time series prediction. The working process of Multiple CNNs is analyzing the periodicity of time series, extracting the closeness and the long and short periodic information of the predicted target respectively, and finally integrating the characteristics of the three parts to make the prediction. Moreover, the model is highly flexible, which allows users to freely adjust the cycle span set in the model according to their own data characteristics. Tests on two large real-world datasets, show that our model has a strong advantage over other time series prediction methods.
Migraine is a neurological disorder characterized by persisting attacks, underlined by the sensitivity to light. One of the leading reasons that make migraine a bigger issue is that it cannot be diagnosed easily by physicians because of the numerous overlapping symptoms with other diseases, such as epilepsy and tension-headache. Consequently, studies have been growing on how to make a computerized decision support system for diagnosis of migraine. In most laboratory studies, flash stimulation is used during the recording of electroencephalogram (EEG) signals with different frequencies and variable (seconds) time windows. The main contribution of this study is the investigation of the effects of flash stimulation on the classification accuracy, and how to find the effective window length for EEG signal classification. To achieve this, we tested different machine learning algorithms on the EEG signals features extracted by using discrete wavelet transform. Our tests on the real-world dataset, recorded in the laboratory, show that the flash stimulation can improve the classification accuracy for more than 10%. Not surprisingly, it is seen that the same holds for the selection of time window length, i.e. the selection of the proper window length is crucial for the accurate migraine identification.
Background:
Brain-computer interface (BCI) is a combination of hardware and software that provides a non-muscular channel to send various messages and commands to the outside world and control external devices such as computers. BCI helps severely disabled patients having neuromuscular injuries, locked-in syndrome (LiS) to lead their life as a normal person to the best extent possible. There are various applications of BCI not only in the field of medicine but also in entertainment, lie detection, gaming, etc. METHODOLOGY: In this work, using BCI a Deceit Identification Test (DIT) is performed based on P300, which has a positive peak from 300 ms to 1000 ms of stimulus onset. The goal is to recognize and classify P300 signals with excellent results. The pre-processing has been performed using the band-pass filter to eliminate the artifacts.
Comparison with existing methods:
Wavelet packet transform (WPT) is applied for feature extraction whereas linear discriminant analysis (LDA) is used as a classifier. Comparison with the other existing methods namely BCD, BAD, BPNN etc has been performed.
Results:
A novel experiment is conducted using EEG acquisition device for the collection of data set on 20 subjects, where 10 subjects acted as guilty and 10 subjects acted as innocent. Training and testing data are in the ratio of 90:10 and the accuracy obtained is up to 91.67%. The proposed approach that uses WPT and LDA results in high accuracy, sensitivity, and specificity.
Conclusion:
The method provided better results in comparison with the other existing methods. It is an efficient approach for deceit identification for EEG based BCI.
The driver's intention is recognized by electroencephalogram(EEG) signals under different driving conditions to provide theoretical and practical support for the applications of automated driving. An EEG signal acquisition system is established by designing a driving simulation experiment, in which data of the driver's EEG signals before turning left, turning right, and going straight, are collected in a specified time window. The collected EEG signals are analyzed and processed by wavelet packet transform to extract characteristic parameters. A driving intention recognition model, based on neural network, is established, and particle swarm optimization (PSO) is adopted to optimize the model parameters. The extracted characteristic parameters are inputted into the recognition model to identify driving intention before turning left, turning right, and going straight. Matlab is used to simulate and verify the established model to obtain the results of the model. The maximum recognition rate of driving intention is 92.9%. Results show that the driver's EEG signal can be used to analyze the law of EEG signals. Furthermore, the PSO-based neural network model can be adapted to recognize driving intention. © 2018 Editorial Department of Journal of Beijing Institute of Technology.
Background: Electroencephalogram (EEG) signals are important for brain health monitoring applications. Characteristics of EEG signals are complex, being non-stationarity, aperiodic and nonlinear in nature. EEG signals are a combination of sustained oscillation and non-oscillation transients that are challenging to deal with using linear approaches.
Method: This research proposes a new scheme based on a tunable Q-factor wavelet transform (TQWT) and a statistical approach to analyse various EEG recordings. Firstly, the proposed method decompose EEG signals into different sub¬¬-bands using the TQWT method, which is parameterized by its Q-factor and redundancy. This method depends on the resonance of a signal, instead of frequency or scaling as in the Fourier and wavelet transforms. Secondly, using a statistical feature extraction on the sub-bands to divide each sub-band into n windows, and then extract several statistical features from each window. Finally, the extracted features are forwarded to a bagging tree (BT), k nearest neighbor (k-NN), and support vector machine (SVM) as classifiers to evaluate the performance of the proposed feature extraction technique.
Results: The proposed method is tested on two different EEG databases: Bonn University database and Born University database. The experimental results demonstrate that the proposed feature extraction algorithm with the k-NN classifier produces the best performance compared with the other two classifiers.
Comparison with existing methods: In order to further evaluate the performances, the proposed scheme is compared with the other existing methods in terms of accuracy. The results prove that the proposed TQWT based feature extraction method has great potential to extract discriminative information from brain signals.
Conclusion: The outcomes of the proposed technique can assist doctors and other health experts to identify diversified EEG categories.
In order to quickly and accurately identify the fatigue related phenomena such as the continuous attention level of the driver after long-time driving and his reaction time to emergencies, the present study establishes a recognition algorithm for recognition of driving-fatigue related problems based on EEG signal index combining the improved particle swarm algorithm with support vector machine, and sets up a method for classifying driving fatigue degree. According to the fatigue classification method, two hours are served as critical points to delineate mild fatigue and severe fatigue. The recognition rate of driver's severe fatigue by the algorithm is higher than that of driver's mild fatigue. The driver's fatigue perception (7.35) in the second phase is much greater than that of the first phase (2.11), and the average reaction time (640 ms) and speed deviation (3.8 km/ h) of the second phase are also much greater than that of the first phase (510 ms) and (1.4 km/ h), indicating that the driver experiences obvious fatigue after driving for two consecutive hours, and his ability to deal with emergencies and to control vehicle during severe fatigue decrease.
Sleep spindles are important components of the N-REM stage-2 in the sleep electroencephalogram (EEG). They are oscillatory EEG activities of fusiform morphology in the range of 10–16 Hz [1], and a duration between 0.5 and 3 s. Spindles have been associated with cognitive skills and sleep-dependent memory consolidation. The aim of this study is to assess differences in the before (“pre”), during (“dur”) and after (“post”) spindle epochs by means of main power spectral bands delta (2–4 Hz), theta (4–8 Hz), alpha (8–12 Hz), beta (12–30 Hz), gamma (30–44 Hz), total (2–44 Hz) and sigma bands (12–16 Hz), calculated by the Welch periodogram, and by Fractal dimension (FD). The analysis was carried out on 7 healthy children (mean age = 8.90 ± 1.34 years) deprived of sleep on the day of the acquisition to enhance the deep sleep during the recording. For each EEG record (standard 10–20, 19 electrodes, sampling rate 512 Hz), two neurophysiologists labeled the start and the end points of the three sleep epochs. The results showed statistical differences between “dur” and both “pre” and “post” epochs in almost all channels (except O1 and O2) for all bands, except gamma. Furthermore, the values of FD were significantly different between “dur” and both “pre” and “post” epochs, for all channels. The FD values in “dur” epochs were smaller than in both “pre” and “post” ones, showing a lower EEG complexity during spindles, compared with the “pre” and “post” epochs. FD values in “post” epochs were found similar to those in “pre” periods. These differences could be useful to comprehend the spindles changes during sleep time. Moreover, these data could help on understanding the system generator of the spindles.
Epilepsy is a neurological disorder and for its detection, encephalography (EEG) is a commonly used clinical approach. Manual inspection of EEG brain signals is a time-consuming and laborious process, which puts heavy burden on neurologists and affects their performance. Several automatic techniques have been proposed using traditional approaches to assist neurologists in detecting binary epilepsy scenarios e.g. seizure vs. non-seizure or normal vs. ictal. These methods do not perform well when classifying ternary case e.g. ictal vs. normal vs. inter-ictal; the maximum accuracy for this case by the state-of-the-art-methods is 97+-1%. To overcome this problem, we propose a system based on deep learning, which is an ensemble of pyramidal one-dimensional convolutional neural network (P-1D-CNN) models. In a CNN model, the bottleneck is the large number of learnable parameters. P-1D-CNN works on the concept of refinement approach and it results in 60% fewer parameters compared to traditional CNN models. Further to overcome the limitations of small amount of data, we proposed augmentation schemes for learning P-1D-CNN model. In almost all the cases concerning epilepsy detection, the proposed system gives an accuracy of 99.1+-0.9% on the University of Bonn dataset.
Driving behavior recognition is the foundation of driver assistance systems, with potential applications in automated driving systems. Most prevailing studies have used subjective questionnaire data and objective driving data to classify driving behaviors, while few studies have used physiological signals such as electroencephalography (EEG) to gather data. To bridge this gap, this paper proposes a two-layer learning method for driving behavior recognition using EEG data. A simulated car-following driving experiment was designed and conducted to simultaneously collect data on the driving behaviors and EEG data of drivers. The proposed learning method consists of two layers. In Layer I, two-dimensional driving behavior features representing driving style and stability were selected and extracted from raw driving behavior data using K-means and support vector machine recursive feature elimination. Five groups of driving behaviors were classified based on these two-dimensional driving behavior features. In Layer II, the classification results from Layer I were utilized as inputs to generate a k-Nearest-Neighbor classifier identifying driving behavior groups using EEG data. Using independent component analysis, a fast Fourier transformation, and linear discriminant analysis sequentially, the raw EEG signals were processed to extract two core EEG features. Classifier performance was enhanced using the adaptive synthetic sampling approach. A leave-one-subject-out cross validation was conducted. The results showed that the average classification accuracy for all tested traffic states was 69.5% and the highest accuracy reached 83.5%, suggesting a significant correlation between EEG patterns and car-following behavior.
Tuberculosis [TB] has afflicted numerous nations in the world. As per a report by the World Health Organization [WHO], an estimated 1.4 million TB deaths in 2015 and an additional 0.4 million deaths resulting from TB disease among people living with HIV, were observed. Most of the TB deaths can be prevented if it is detected at an early stage. The existing processes of diagnosis like blood tests or sputum tests are not only tedious but also take a long time for analysis and cannot differentiate between different drug resistant stages of TB. The need to find newer prompt methods for disease detection has been aided by the latest Artificial Intelligence [AI] tools. Artificial Neural Network [ANN] is one of the important tools that is being used widely in diagnosis and evaluation of medical conditions. This review aims at providing brief introduction to various AI tools that are used in TB detection and gives a detailed description about the utilization of ANN as an efficient diagnostic technique. The paper also provides a critical assessment of ANN and the existing techniques for their diagnosis of TB. Researchers and Practitioners in the field are looking forward to use ANN and other upcoming AI tools such as Fuzzy-logic, genetic algorithms and artificial intelligence simulation as a promising current and future technology tools towards tackling the global menace of Tuberculosis. Latest advancements in the diagnostic field include the combined use of ANN with various other AI tools like the Fuzzy-logic, which has led to an increase in the efficacy and specificity of the diagnostic techniques.
With the advances in pervasive sensor technologies, physiological signals can be captured continuously to prevent the serious outcomes caused by epilepsy. Detection of epileptic seizure onset on collected multi-channel electroencephalogram (EEG) has attracted lots of attention recently. Deep learning is a promising method to analyze large-scale unlabeled data. In this paper, we propose a multi-view deep learning model to capture brain abnormality from multi-channel epileptic EEG signals for seizure detection. Specifically, we first generate EEG spectrograms using short-time Fourier transform (STFT) to represent the time-frequency information after signal segmentation. Second, we adopt stacked sparse denoising autoencoders (SSDA) to unsupervisedly learn multiple features by considering both intra and inter correlation of EEG channels, denoted as intra-channel and cross-channel features, respectively. Third, we add an SSDA-based channel selection procedure using proposed response rate to reduce the dimension of intra-channel feature. Finally, we concatenate the learned multi-features and apply a fully-connected SSDA model with softmax classifier to jointly learn the cross-patient seizure detector in a supervised fashion. To evaluate the performance of the proposed model, we carry out experiments on a real world benchmark EEG dataset and compare it with six baselines. Extensive experimental results demonstrate that the proposed learning model is able to extract latent features with meaningful interpretation, and hence is effective in detecting epileptic seizure.
The traditional E-learning system is limited with monitoring attention level of students. The online instructor cannot monitor whether the students remain focus during online autonomous learning. Along with the attention, emotions are also intrinsically related to the way that individuals interact with each other as well as machines. The behavior and emotions can be better understood by a human being to improve the communication but a machine cannot. This gives rise to lack of interest and knowledge gain. To overcome this limitation of the traditional E-learning system EEG sensors are been introduced. Using, EEG we can detect the emotion of the user, by actually looking inside the users brain to check user's mental state and signal waves are generated accordingly as the output which can be used to improve users learning experience. The proposed system is a smart E-learning system which predicts the video based on emotion. The system uses Neurosky Brainwave detector and Random forest Classification method to classify the waves to predict appropriate emotions.
Indicator dilution theory predicts that the first-pass pulmonary and systemic arterial concentrations of a drug will be inversely related to the cardiac output. For high-clearance drugs, these first-pass concentrations may contribute significantly to the measured arterial concentrations, which would therefore also be inversely related to cardiac output. We examined the cardiac output dependence of the initial kinetics of propofol in two separate studies using chronically instrumented sheep in which propofol (100 mg) was infused IV over 2 min. In the first study, steady-state periods of low, medium, and high cardiac output were achieved by altering carbon dioxide tension in six halothane-anesthetized sheep. The initial area under the curve and peak value of the pulmonary artery propofol concentrations were inversely related to cardiac output (R² = 0.57 and 0.66, respectively). For the systemic arterial concentrations, these R² values were 0.68 and 0.71, respectively. In our second study, transient reductions in cardiac output were achieved in five conscious sheep by administering a short infusion of metaraminol concurrently with propofol. Cardiac output was lowered by 2.2 L/min, and the area under the curve to 10 min of the arterial concentrations increased to 143% of control.
Implications: The initial arterial concentrations of propofol after IV administration were shown to be inversely related to cardiac output. This implies that cardiac output may be a determinant of the induction of anesthesia with propofol.
Background:
Multi-Scale Ranked Organizing Map coupled with Implicit Function as Squashing Time algorithm(MS-ROM/I-FAST) is a new, complex system based on Artificial Neural networks (ANNs) able to extract features of interest in computerized EEG through the analysis of few minutes of their EEG without any preliminary pre-processing. A proof of concept study previously published showed accuracy values ranging from 94%-98% in discerning subjects with Mild Cognitive Impairment and/or Alzheimer's Disease from healthy elderly people. The presence of deviant patterns in simple resting state EEG recordings in autism, consistent with the atypical organization of the cerebral cortex present, prompted us in applying this potent analytical systems in search of a EEG signature of the disease.
Aim of the study:
The aim of the study is to assess how effectively this methodology distinguishes subjects with autism from typically developing ones.
Methods:
Fifteen definite ASD subjects (13 males; 2 females; age range 7-14; mean value = 10.4) and ten typically developing subjects (4 males; 6 females; age range 7-12; mean value 9.2) were included in the study. Patients received Autism diagnoses according to DSM-V criteria, subsequently confirmed by the ADOS scale. A segment of artefact-free EEG lasting 60 seconds was used to compute input values for subsequent analyses. MS-ROM/I-FAST coupled with a well-documented evolutionary system able to select predictive features (TWIST) created an invariant features vector input of EEG on which supervised machine learning systems acted as blind classifiers.
Results:
The overall predictive capability of machine learning system in sorting out autistic cases from normal control amounted consistently to 100% with all kind of systems employed using training-testing protocol and to 84% - 92.8% using Leave One Out protocol. The similarities among the ANN weight matrixes measured with apposite algorithms were not affected by the age of the subjects. This suggests that the ANNs do not read age-related EEG patterns, but rather invariant features related to the brain's underlying disconnection signature.
Conclusion:
This pilot study seems to open up new avenues for the development of non-invasive diagnostic testing for the early detection of ASD.
In this issue of JAMA, Gulshan and colleagues⁵ present findings from a study evaluating the use of deep learning for detection of diabetic retinopathy and macular edema. To build their model, the authors collected 128 175 annotated images from the EyePACs database. Each image was rated by 3 to 7 clinicians for referable diabetic retinopathy, diabetic macular edema, and overall image quality. Each rater was selected from a panel of 54 board-certified ophthalmologists and senior ophthalmology residents. Using this data set, the algorithm learned to predict the consensus grade of the raters along each clinical attribute: referable diabetic retinopathy, diabetic macular edema, and image quality. To validate their algorithm, the authors assessed its performance on 2 separate and nonoverlapping data sets consisting of 9963 and 1748 images. On the validation data, the algorithm had high sensitivity and specificity. Only one of these values (sensitivity on the second validation data set) failed to be superior at a statistically significant level. The other performance metrics (eg, area under the receiver operating characteristic curve, negative predictive value, positive predictive value) were likewise impressive, giving the authors confidence that this algorithm could be of clinical utility.
The dominant paradigm for video-based action segmentation is composed of two steps: first, for each frame, compute low-level features using Dense Trajectories or a Convolutional Neural Network that encode spatiotemporal information locally, and second, input these features into a classifier that captures high-level temporal relationships, such as a Recurrent Neural Network (RNN). While often effective, this decoupling requires specifying two separate models, each with their own complexities, and prevents capturing more nuanced long-range spatiotemporal relationships. We propose a unified approach, as demonstrated by our Temporal Convolutional Network (TCN), that hierarchically captures relationships at low-, intermediate-, and high-level time-scales. Our model achieves superior or competitive performance using video or sensor data on three public action segmentation datasets and can be trained in a fraction of the time it takes to train an RNN.
The aim of this study is to design a robust feature extraction method for the classification of multiclass EEG signals to determine valuable features from original epileptic EEG data and to discover an efficient classifier for the features. An optimum allocation based principal component analysis method named as OA_PCA is developed for the feature extraction from epileptic EEG data. As EEG data from different channels are correlated and huge in number, the optimum allocation (OA) scheme is used to discover the most favorable representatives with minimal variability from a large number of EEG data. The principal component analysis (PCA) is applied to construct uncorrelated components and also to reduce the dimensionality of the OA samples for an enhanced recognition. In order to choose a suitable classifier for the OA_PCA feature set, four popular classifiers: least square support vector machine (LS-SVM), naive bayes classifier (NB), k-nearest neighbor algorithm (KNN), and linear discriminant analysis (LDA) are applied and tested. Furthermore, our approaches are also compared with some recent research work. The experimental results show that the LS-SVM_1v1 approach yields 100% of the overall classification accuracy (OCA), improving up to 7.10% over the existing algorithms for the epileptic EEG data. The major finding of this research is that the LS-SVM with the 1v1 system is the best technique for the OA_PCA features in the epileptic EEG signal classification that outperforms all the recent reported existing methods in the literature.
Copyright © 2015 Elsevier Ireland Ltd. All rights reserved.
Covariate Model selection for population PK/PD models represents a daunting task because of the large variety of possible alternative covariates that can enter a structural model, the different models that can express the relationship parameter/covariates, and the number of alternative models that can be considered. After describing the problem and briefly reviewing the past literature dedicated to the solution of the problem we use simulations to show the limitations of current approaches and propose an alternative based on the sequential use of Bayesian Trans Dimensional Models. Although the alternative mollifies the dimensionality problem associated with covariate selection, we argue that the overall approach to covariate modeling within PKPD models might need to be reconsidered.
Big Data has become one of the most commonly used terms in the Information Technology circles. The sheer volume of data to be processed and analyzed has grown exponentially with the increasing popularity of Internet and World Wide Web. This presents challenges while storing, manipulating and mining the data. Every day researchers are working on different solutions to handle the volume of data being provided. In machine learning, classification of new observations is done on the basis of the provided learning(training) data to the classifiers. One of the most commonly used efficient and accurate classifiers is the Naive Bayesian classifier. This paper proposes a novel parallel implementation of Naive Bayesian (PNB+) classifier to decrease the testing time complexity while handling large data sets.
