Leave-one-subject-out Research Articles

CubicPat: Investigations on the Mental Performance and Stress Detection Using EEG Signals

Background\Objectives: Solving the secrets of the brain is a significant challenge for researchers. This work aims to contribute to this area by presenting a new explainable feature engineering (XFE) architecture designed to obtain explainable results related to stress and mental performance using electroencephalography (EEG) signals. Materials and Methods: Two EEG datasets were collected to detect mental performance and stress. To achieve classification and explainable results, a new XFE model was developed, incorporating a novel feature extraction function called Cubic Pattern (CubicPat), which generates a three-dimensional feature vector by coding channels. Classification results were obtained using the cumulative weighted iterative neighborhood component analysis (CWINCA) feature selector and the t-algorithm-based k-nearest neighbors (tkNN) classifier. Additionally, explainable results were generated using the CWINCA selector and Directed Lobish (DLob). Results: The CubicPat-based model demonstrated both classification and interpretability. Using 10-fold cross-validation (CV) and leave-one-subject-out (LOSO) CV, the introduced CubicPat-driven model achieved over 95% and 75% classification accuracies, respectively, for both datasets. Conclusions: The interpretable results were obtained by deploying DLob and statistical analysis.

Alzheimer’s Disease Diagnosis from EEG Signals using Recurrence Qualification Analysis (RQA) Features

Alzheimer’s disease (AD) is a neurological disorder that causes memory decline and loss of cognitive abilities. AD directly impacts the brain activity of affected individuals, which can be reflected in electroencephalogram (EEG) signals. Previous studies have typically relied on statistical, spectral, and wavelet features in order detect AD using EEG signals. Recurrence Qualification Analysis (RQA) is a non-linear technique that has been successfully used to analyze EEG signals in several domains including emotion recognition and autism detection. However, RQA features have not yet been fully investigated for AD diagnosis. The aim of this work is to thoroughly investigate the usefulness of RQA features for AD diagnosis. Fifteen RQA features were computed along with statistical measures, Hjorth parameters, and relative power in order to compare the performance of the RQA features to other commonly utilized EEG features. All features were computed from the different brain regions. Experimental results indicated that RQA features outperformed all other feature types regardless of the considered brain region. RQA features achieved accuracies ranging from 89.6% to 98.0% using a support machine vector (SVM) classifier with leave-one-subject-out (LOSO) cross-validation. These results are between 25% to 40% higher than the three other considered feature groups. Feature ranking was performed to find the most relevant RQA features, identifying seven of the fifteen computed RQA features. This work sheds light on the potential of RQA features for reliable AD diagnosis and paves the way for their integration in computer-aided AD diagnostic tools using EEG signals.

Identification of Balance biomarkers using Inertial sensor for Early Detection of Dementia

AbstractBackgroundEarly‐stage dementia, Mild Cognitive Impairment (MCI), is challenging to diagnose since it is a transient condition distinct from complete cognitive collapse. Recent clinical research studies have identified that balance impairments can be a significant indicator for predicting dementia in older adults. Accordingly, we aimed to identify key balance biomarkers using wearable inertial sensors for early detection of dementia/MCI.MethodAt National Research Center for Dementia (South Korea), 60 participants (30 Cognitively Normal‐CN, 30 Mild Cognitive Impairment‐MCI) were selected based on assessments conducted by medical professionals at Chosun University Hospital / Chonnam National University Hospital (Gwangju). Shimmer‐3 inertial sensor was placed on lower back of participants, with a tri‐axial accelerometer and tri‐axial gyroscope. The data was collected under four different conditions: eyes‐open (EO), eyes‐closed (EC), right‐leg lift (RL), left‐leg lift (LL). A set of 76 postural sway measures were considered with 43 time‐domain and 33 frequency‐domain features.Feature ranking was executed through the utilization of "Leave‐One‐Subject‐Out (LOSO)" cross‐validation technique for each condition. Filter methods: one‐way‐ANOVA (Analysis of Variance), Mutual Information and Wrapper methods: Random Forest, Support Vector Machine (SVM) were utilized to evaluate feature subsets. The proportional value of each feature was calculated by the number of times it showed in the top feature list. The final score was obtained by combining scores from all the techniques, which indicated the relative weight of characteristics across the various approaches. Further feature analysis was conducted using SHAP (Shapley‐Additive‐exPlanations).ResultThe multi‐step methodology, for feature selection, adopted in our research ensured a rigorous process, combining the strengths of Filter and Wrapper methods. In time‐domain: root‐mean‐square, mean‐distance, 95%‐confidence circle sway‐area, path‐length, mean sway‐velocity were the key features identified. While in frequency‐domain: spectral edge‐frequency, mean‐frequency, peak‐frequency, centroid‐frequency, spectral‐entropy were found to be the principal features of relevance. These features were also included in the top list of features based on SHAP impact value, hence proving the effectiveness of our method in identifying important biomarkers for detection of MCI.ConclusionThe inertial sensors based postural balance assessment provides useful biomarkers that can facilitate early screening of MCI / dementia in home settings.

QuadTPat: Quadruple Transition Pattern-based explainable feature engineering model for stress detection using EEG signals

The most cost-effective data collection method is electroencephalography (EEG), which obtains meaningful information about the brain. Therefore, EEG signal processing is crucial for neuroscience and machine learning (ML). Therefore, a new EEG stress dataset has been collected, and an explainable feature engineering (XFE) model has been proposed using the Directed Lobish (DLob) symbolic language. The first phase of this research is the data collection phase, and an EEG stress dataset was gathered from 310 participants. This collected stress dataset contains two classes: (i) stress and (ii) control. An XFE model has been presented to detect stress automatically. The presented XFE model has four main phases, and these are (i) channel transformer and quadruple transition pattern (QuadTPat)-based feature generation, (ii) feature selection deploying cumulative weighted neighborhood component analysis (CWNCA), (iii) explainable results creation with DLob and (iv) classification with t algorithm-based k-nearest neighbors (tkNN) classifier. The proposed XFE model generates a DLob string, and the explainable results were obtained using this string. Moreover, the presented XFE model attained 92.95% and 73.63% classification accuracy, deploying 10-fold and leave-one subject-out (LOSO) cross-validations (CV). According to the classification performances, the recommended QuadTPat-based XFE is a good model for EEG signal classification. Also, the presented QuadTPat-based XFE model is a good model for explainable artificial intelligence (XAI) since TTPat-based XFE is cooperating with the DLob.

TATPat based explainable EEG model for neonatal seizure detection

The most cost-effective data collection method is electroencephalography (EEG) to obtain meaningful information about the brain. Therefore, EEG signal processing is very important for neuroscience and machine learning (ML). The primary objective of this research is to detect neonatal seizures and explain these seizures using the new version of Directed Lobish. This research uses a publicly available neonatal EEG signal dataset to get comparative results. In order to classify these EEG signals, an explainable feature engineering (EFE) model has been proposed. In this EFE model, there are four essential phases and these phases: (i) automaton and transformer-based feature extraction, (ii) feature selection deploying cumulative weight-based neighborhood component analysis (CWNCA), (iii) the Directed Lobish (DLob) and Causal Connectome Theory (CCT)-based explainable result generation and (iv) classification deploying t algorithm-based support vector machine (tSVM). In the first phase, we have used a channel transformer to get channel numbers and these values have been divided into three levels and these levels are named (1) high, (2) medium and (3) low. By utilizing these levels, we have created an automaton and this automaton has three nodes (each node defines each level). In the feature extraction phase, transition tables of these nodes has been extracted. Therefore, the proposed feature extraction function is termed Triple Nodes Automaton-based Transition table Pattern (TATPat). The used EEG signal dataset contains 19 channels and there are 9 (= 32) connection in the defined automaton. Thus, the presented TATPat extracts 3249 (= 19 × 19 × 9) features from each EEG segment. To choose the most informative features of these 3249 features, a new feature selector which is CWNCA has been applied. By cooperating findings of this feature selector and the presented DLob, the explainable results have been obtained. The last phase is the classification phase and to get high classification performance from this phase, an ensemble classifier (tSVM) has been presented and the classification results have been obtained using two validation techniques which are 10-fold cross-validation (CV) and leave-one subject-out (LOSO) CV. The proposed EFE model generates a DLob string and by using this string, the explainable results have been obtained. Moreover, the presented EFE model attained 99.15% and 76.37% classification accuracy deploying 10-fold and LOSO CVs respectively. According to the classification performances, the recommended TATPat-based EFE is a good model at EEG signal classification. Also, the presented TATPat-based EFE model is a good model for explainable artificial intelligence (XAI) since TTPat-based EFE is cooperating by the DLob.

Compact convolutional transformer for subject-independent motor imagery EEG-based BCIs

Motor imagery electroencephalography (EEG) analysis is crucial for the development of effective brain-computer interfaces (BCIs), yet it presents considerable challenges due to the complexity of the data and inter-subject variability. This paper introduces EEGCCT, an application of compact convolutional transformers designed specifically to improve the analysis of motor imagery tasks in EEG. Unlike traditional approaches, EEGCCT model significantly enhances generalization from limited data, effectively addressing a common limitation in EEG datasets. We validate and test our models using the open-source BCI Competition IV datasets 2a and 2b, employing a Leave-One-Subject-Out (LOSO) strategy to ensure subject-independent performance. Our findings demonstrate that EEGCCT not only outperforms conventional models like EEGNet in standard evaluations but also achieves better performance compared to other advanced models such as Conformer, Hybrid s-CViT, and Hybrid t-CViT, while utilizing fewer parameters and achieving an accuracy of 70.12%. Additionally, the paper presents a comprehensive ablation study that includes targeted data augmentation, hyperparameter optimization, and architectural improvements.

Automated EEG-based language detection using directed quantum pattern technique

Electroencephalogram (EEG) signals contain complex useful information about brain activities. These EEG signals are noisy, highly varying and nonstationary in nature. Hence, extracting meaningful information from these signals is challenging. The existing machine learning systems struggle to capture the minute changes from the signals and yield high performance.This study introduces a novel quantum-inspired feature extraction technique called Directed Quantum Pattern (DQP), designed to address these challenges by using a lattice structure to capture directional binary features. These directions (paths) are computed using a maximum function providing a dynamic and adaptive feature representation.This paper presents a novel DQP-LangNet developed using DQP for automated classification of two- languages using EEG signals. We have proposed a hybrid approach, combining DQP, statistical features, and multi-level discrete wavelet transform (MDWT) to extract salient features similar to the deep learning approach. The EEG dataset consisting of 14 channels, produces 7 feature vectors per channel, yielding 98 feature vectors. Neighborhood component analysis and Chi-square (Chi2) feature selection approaches generated 196 feature vectors.In addition to the innovative feature extraction a new classification structure called “t” is proposed k-nearest neighbor (tkNN) and support vector machine (tSVM) classifiers are employed. Using the proposed tkNN and tSVM classifiers, 392 (=196×2) classifier-based outcomes are obtained. To further improve classification performance, we applied the iterative majority voting (IMV) technique to automatically select the best result.Our DQP-based model achieved a classification accuracy of 95.68 %using EEG language dataset with leave-one-subject-out (LOSO) cross-validation strategy. Also, an explainable feature engineering (XFE) structure of DQP-LangNet is employed to obtain channel-specific explainable results. Our proposed DQP-LangNet model can be employed for other applications in neuroscience.

Emotion recognition using cross-modal attention from EEG and facial expression

The fusion of facial expression and Electroencephalogram (EEG) signals in a multi-modal framework is a comprehensive and accurate method for emotion recognition. However, current methods often tend to directly concatenate two modalities, thereby overlooking the interplay between modalities. Furthermore, the selection of single static facial images for extracting facial features neglected the dynamic changes in facial expressions. These limitations have been identified as factors that constrain the accuracy and stability of the model. In this study, an innovative multimodal emotion recognition network was introduced, which using continuous facial expressions and EEG signals. It incorporated the cross-modal attention fusion mechanism to establish robust correlations between modal feature vectors, thereby generating amalgamated vectors with mutual information. Additionally, a Self-Attention Convolutional Long Short-Term Memory (SA-ConvLSTM) was employed to capture spatiotemporal information from facial expression images. The model proposed in this paper was experimentally evaluated on the DEAP and MAHNOB-HCI datasets. Its recognition accuracy was higher than the existing advanced research methods. At the same time, it still performed well in the Leave-One-Subject-Out (LOSO) experiment based on DEAP dataset. The experimental results showed the effectiveness of the proposed model in the task of multimodal emotion recognition.

A generalized thermal comfort model using thermographic images and compact convolutional transformers: Towards scalable and adaptive occupant comfort optimization

Thermal comfort models that account for individual thermal sensations are crucial for optimizing environmental control while maintaining occupant comfort. However, the current approach of developing occupant-specific models is not scalable for new occupants whose data were not used for training, making it impractical for multi-occupant spaces and burdensome for management. To address this, we propose a thermal comfort model based on a subject-independent evaluation approach, capable of generalizing to new individuals not included during model training. This model accurately predicts thermal sensations without the need for occupant-specific models, allowing scalability and adaptability to new occupants. Furthermore, this study considers occupants with face masks, which is essential in environments where masks are required for their wellbeing and performance. The model was developed using Compact Convolutional Transformers which combines convolutional layers and transformer layers, allowing the model to capture both fine-grained local features and broader contextual relationships, making it effective for predicting thermal comfort from thermal images. The model was trained using the Leave-One-Subject-Out (LOSO) and Leave-One-Group-Out (LOGO) training approaches. The LOSO approach achieved high accuracies of 98.04 %, 98.93 %, and 98.88 % for dataset A (participants without face masks), dataset B (participants with face masks), and dataset C (a combination of both), respectively. The LOGO approach achieved accuracies of 99.92 %, 99.40 %, and 99.85 % for datasets A, B, and C, respectively, showing a slightly better prediction performance. The approach presented in this study offers a promising solution for designing accurate generalized models for real-time environmental control to meet the thermal comfort needs of occupants.

Cognitive activity analysis of Parkinson's patients using artificial intelligence techniques.

The development of modern Artificial Intelligence (AI) based models for the early diagnosis of Parkinson's disease (PD) has been gaining deep attention by researchers recently. In particular, the use of different types of datasets (voice, hand movements, gait, etc.) increases the variety of up-to-date models. Movement disorders and tremors are also among the most prominent symptoms of PD. The usage of drawings in the detection of PD can be a crucial decision-support approach that doctors can benefit from. A dataset was created by asking 40 PD and 40 Healthy Controls (HC) to draw spirals with and without templates using a special tablet. The patient-healthy distinction was achieved by classifying drawings of individuals using Support Vector Machine (SVM), Random Forest (RF), and Naive Bayes (NB) algorithms. Prior to classification, the data were normalized by applying the min-max normalization method. Moreover, Leave-One-Subject-Out (LOSO) Cross-Validation (CV) approach was utilized to eliminate possible overfitting scenarios. To further improve the performances of classifiers, Principal Component Analysis (PCA) dimension reduction technique were also applied to the raw data and the results were compared accordingly. The highest accuracy among machine learning based classifiers was obtained as 90% with SVM classifier using non-template drawings with PCA application. The model can be used as a pre-evaluation system in the clinic as a non-invasive method that also minimizes environmental and educational level differences by using simple hand gestures such as hand drawing, writing numbers, words, and syllables. As a result of our study, preliminary preparation has been made so that hand drawing analysis can be used as an auxiliary system that can save time for health professionals. We plan to work on more comprehensive data in the future.

Human activity recognition system using wearable accelerometers for classification of leg movements: a first, detailed approach

Abstract. A human activity recognition (HAR) system carried by masseurs for controlling a therapy table via different movements of legs or hip is studied. This work starts with a survey on HAR systems using the sensor position named “trouser pockets”. Afterwards, in the experiments, the impacts of different hardware systems, numbers of subjects, data generation processes (online streams/offline data snippets), sensor positions, sampling rates, sliding window sizes and shifts, feature sets, feature elimination processes, operating legs, tag orientations, classification processes (concerning method, parameters and an additional smoothing process), numbers of activities, training databases, and the use of a preceding teaching process on the classification accuracy are examined to get a thorough understanding of the variables influencing the classification quality. Besides the impacts of different adjustable parameters, this study also serves as an advisor for the implementation of classification tasks. The proposed system has three operating classes: do nothing, pump therapy table up or pump therapy table down. The first operating class consists of three activity classes (go, run, massage) such that the whole classification process exists with five classes. Finally, using online data streams, a classification accuracy of 98 % could be achieved for one skilled subject and about 90 % for one randomly chosen subject (mean of 1 skilled and 11 unskilled subjects). With the LOSO (leave-one-subject-out) technique for 12 subjects, up to 86 % can be attained. With our offline data approach, we get accuracies of 98 % for 12 subjects and up to 100 % for 1 skilled subject.

A non-invasive AI-based system for precise grading of anosmia in COVID-19 using neuroimaging

COVID-19 (Coronavirus), an acute respiratory disorder, is caused by SARS-CoV-2 (coronavirus severe acute respiratory syndrome). The high prevalence of COVID-19 infection has drawn attention to a frequent illness symptom: olfactory and gustatory dysfunction. The primary purpose of this manuscript is to create a Computer-Assisted Diagnostic (CAD) system to determine whether a COVID-19 patient has normal, mild, or severe anosmia. To achieve this goal, we used fluid-attenuated inversion recovery (FLAIR) Magnetic Resonance Imaging (FLAIR-MRI) and Diffusion Tensor Imaging (DTI) to extract the appearance, morphological, and diffusivity markers from the olfactory nerve. The proposed system begins with the identification of the olfactory nerve, which is performed by a skilled expert or radiologist. It then proceeds to carry out the subsequent primary steps: (i) extract appearance markers (i.e., 1st and 2nd order markers), morphology/shape markers (i.e., spherical harmonics), and diffusivity markers (i.e., Fractional Anisotropy (FA) & Mean Diffusivity (MD)), (ii) apply markers fusion based on the integrated markers, and (iii) determine the decision and corresponding performance metrics based on the most-promising classifier. The current study is unusual in that it ensemble bags the learned and fine-tuned ML classifiers and diagnoses olfactory bulb (OB) anosmia using majority voting. In the 5-fold approach, it achieved an accuracy of 94.1%, a balanced accuracy (BAC) of 92.18%, precision of 91.6%, recall of 90.61%, specificity of 93.75%, F1 score of 89.82%, and Intersection over Union (IoU) of 82.62%. In the 10-fold approach, stacking continued to demonstrate impressive results with an accuracy of 94.43%, BAC of 93.0%, precision of 92.03%, recall of 91.39%, specificity of 94.61%, F1 score of 91.23%, and IoU of 84.56%. In the leave-one-subject-out (LOSO) approach, the model continues to exhibit notable outcomes, achieving an accuracy of 91.6%, BAC of 90.27%, precision of 88.55%, recall of 87.96%, specificity of 92.59%, F1 score of 87.94%, and IoU of 78.69%. These results indicate that stacking and majority voting are crucial components of the CAD system, contributing significantly to the overall performance improvements. The proposed technology can help doctors assess which patients need more intensive clinical care.

Selecting EEG channels and features using multi-objective optimization for accurate MCI detection: validation using leave-one-subject-out strategy

Effective management of dementia requires the timely detection of mild cognitive impairment (MCI). This paper introduces a multi-objective optimization approach for selecting EEG channels (and features) for the purpose of detecting MCI. Firstly, each EEG signal from each channel is decomposed into subbands using either variational mode decomposition (VMD) or discrete wavelet transform (DWT). A feature is then extracted from each subband using one of the following measures: standard deviation, interquartile range, band power, Teager energy, Katz's and Higuchi's fractal dimensions, Shannon entropy, sure entropy, or threshold entropy. Different machine learning techniques are used to classify the features of MCI cases from those of healthy controls. The classifier's performance is validated using leave-one-subject-out (LOSO) cross-validation (CV). The non-dominated sorting genetic algorithm (NSGA)-II is designed with the aim of minimizing the number of EEG channels (or features) and maximizing classification accuracy. The performance is evaluated using a publicly available online dataset containing EEGs from 19 channels recorded from 24 participants. The results demonstrate a significant improvement in performance when utilizing the NSGA-II algorithm. By selecting only a few appropriate EEG channels, the LOSO CV-based results show a significant improvement compared to using all 19 channels. Additionally, the outcomes indicate that accuracy can be further improved by selecting suitable features from different channels. For instance, by combining VMD and Teager energy, the SVM accuracy obtained using all channels is 74.24%. Interestingly, when only five channels are selected using NSGA-II, the accuracy increases to 91.56%. The accuracy is further improved to 95.28% when using only 8 features selected from 7 channels. This demonstrates that by choosing informative features or channels while excluding noisy or irrelevant information, the impact of noise is reduced, resulting in improved accuracy. These promising findings indicate that, with a limited number of channels and features, accurate diagnosis of MCI is achievable, which opens the door for its application in clinical practice.

EEG-based schizophrenia detection using fusion of effective connectivity maps and convolutional neural networks with transfer learning.

Schizophrenia (SZ) is a serious mental disorder that can mainly be distinguished by symptoms including delusions and hallucinations. This mental disorder makes difficult conditions for the person and her/his relatives. Electroencephalogram (EEG) signal is a sophisticated neuroimaging technique that helps neurologists to diagnose this mental disorder. Estimating and evaluating brain effective connectivity between electrode pairs is an appropriate way of diagnosing brain states in neuroscience studies. In this study, we construct a novel image from multi-channels of EEG based on the fusion of three effective connectivity, partial directed coherence (PDC), and direct directed transfer function (dDTF) and transfer entropy (TE) at three consecutive time windows. Then, this image was used as input of five well-known convolutional neural networks (CNNs) through transfer learning (TL) to learn patterns related to SZ patients to diagnose this disorder from normal participants from two public databases. Also, the majority voting method was used to improve these results based on ensemble results of the five CNNs, i.e., ResNet-50, Inception-v3, DenseNet-201, EfficientNetB0, and NasNet-Mobile. The highest average accuracy, specificity and sensitivity to diagnose SZ patients from healthy participants were obtained using EfficientNetB0 through the Leave-One-Subject-out (LOSO) Cross-Validation criterion equal to 96.67%, 96.23%, 96.82%, 95.15%, 94.42% and 96.28% for the first and second databases, respectively. Also, as we suggested, the ensemble approach of EfficientNetB0, ResNet-50 and NasNet-Mobile increased the accuracy by approximately 3%. Our results show the effectiveness of providing fused images from multichannel EEG signals to the ensemble of CNNs through TL to diagnose SZ than state-of-the-art studies.

A 4D spontaneous micro‐expression database: Establishment and evaluation

AbstractMicro‐expressions are spontaneous and unconscious facial movements that reveal individuals’ genuine inner emotions. They hold significant potential in various psychological testing fields. As the face is a 3D deformation object, the emergence of facial expression leads to spatial deformation of the face. However, existing databases primarily offer 2D video sequences, limiting descriptions of 3D spatial information related to micro‐expressions. Here, a new micro‐expression database is proposed, which contains 2D image sequences and corresponding 3D point cloud sequences. These samples were classified using both an objective method based on the facial action coding system and a non‐objective emotion classification method that considers video contents and participants’ self‐reports. A variety of feature extraction techniques are applied to 2D data, including traditional algorithms and deep learning methods. Additionally, a novel local curvature‐based algorithm is developed to extract 3D spatio‐temporal deformation features from the 3D data. The authors evaluated the classification accuracies of these two features individually and their fusion results under leave‐one‐subject‐out (LOSO) and tenfold cross‐validation. The results demonstrate that fusing 3D features with 2D features results in improved recognition performance compared to using 2D features alone.

Prediction of central nervous system oxygen toxicity symptoms using electrodermal activity and machine learning

ObjectiveBreathing elevated oxygen partial pressures (PO2) prior to SCUBA diving increases the risk of developing central nervous system oxygen toxicity (CNS-OT), which could impair performance or result in seizure and subsequent drowning. We aimed to study the dynamics of electrodermal activity (EDA) while breathing elevated PO2 in the hyperbaric environment (HBO2) as a possible means to predict impending CNS-OT. To this end, we used machine learning to automatically detect and predict the onset of symptoms associated with CNS-OT in humans by using features derived from EDA in both time and frequency domains. MethodsWe collected electrodermal activity (EDA) data from forty-nine exposures to HBO2 while subjects were undergoing cognitive load and exercise in a hyperbaric oxygen chamber. Four independent experts were present during the experiment to monitor and classify any symptoms associated with hyperbaric oxygen toxicity. We computed a highly sensitive time varying spectral EDA index, named TVSymp, and extracted informative features from skin conductance responses (SCRs). Machine learning algorithms were trained and validated for classifying features from SCRs and TVSymp as CNS-OT related or non-CNS-OT related. Machine learning models were validated using a subject-independent leave one subject out (LOSO) validation scheme. ResultsOur machine learning model was able to classify EDA dynamics related to CNS-OT with 100 % sensitivity and 84 % specificity via LOSO validation. Moreover, the median prediction time for CNS-OT symptoms was ∼ 250 s preceding the occurrence of actual symptoms. SignificanceThis study shows that EDA can potentially be used for early prediction of CNS-OT in divers with a high sensitivity and sufficient prediction time for countermeasures. While the study results are promising, independent validation datasets are warranted to confirm the findings. However, the current results are well corroborated in an animal study, which consistently showed seizure prediction time of 2 min prior to seizure.

Feature evaluation of accelerometry signals for cough detection.

Cough is a common symptom of multiple respiratory diseases, such as asthma and chronic obstructive pulmonary disorder. Various research works targeted cough detection as a means for continuous monitoring of these respiratory health conditions. This has been mainly achieved using sophisticated machine learning or deep learning algorithms fed with audio recordings. In this work, we explore the use of an alternative detection method, since audio can generate privacy and security concerns related to the use of always-on microphones. This study proposes the use of a non-contact tri-axial accelerometer for motion detection to differentiate between cough and non-cough events/movements. A total of 43 time-domain features were extracted from the acquired tri-axial accelerometry signals. These features were evaluated and ranked for their importance using six methods with adjustable conditions, resulting in a total of 11 feature rankings. The ranking methods included model-based feature importance algorithms, first principal component, leave-one-out, permutation, and recursive features elimination (RFE). The ranking results were further used in the feature selection of the top 10, 20, and 30 for use in cough detection. A total of 68 classification models using a simple logistic regression classifier are reported, using two approaches for data splitting: subject-record-split and leave-one-subject-out (LOSO). The best-performing model out of the 34 using subject-record-split obtained an accuracy of 92.20%, sensitivity of 90.87%, specificity of 93.52%, and F1 score of 92.09% using only 20 features selected by the RFE method. The best-performing model out of the 34 using LOSO obtained an accuracy of 89.57%, sensitivity of 85.71%, specificity of 93.43%, and F1 score of 88.72% using only 10 features selected by the RFE method. These results demonstrate the ability for future implementation of a motion-based wearable cough detector.

A novel in-bed body posture monitoring for decubitus ulcer prevention using body pressure distribution mapping

BackgroundDecubitus ulcers are prevalent among the aging population due to a gradual decline in their overall health, such as nutrition, mental health, and mobility, resulting in injury to the skin and tissue. The most common technique to prevent these ulcers is through frequent repositioning to redistribute body pressures. Therefore, the main goal of this study is to facilitate the timely repositioning of patients through the use of a pressure mat to identify in-bed postures in various sleep environments. Pressure data were collected from 10 healthy participants lying down on a pressure mat in 19 various in-bed postures, correlating to the supine, prone, right-side, and left-side classes. In addition, pressure data were collected from participants sitting at the edge of the bed as well as an empty bed. Each participant was asked to lie in these 19 postures in three distinct testing environments: a hospital bed, a home bed, and a home bed with a foam mattress topper. To categorize each posture into its respective class, the pre-trained 2D ResNet-18 CNN and the pre-trained Inflated 3D CNN algorithms were trained and validated using image and video pressure mapped data, respectively.ResultsThe ResNet-18 and Inflated 3D CNN algorithms were validated using leave-one-subject-out (LOSO) and leave-one-environment-out (LOEO) cross-validation techniques. LOSO provided an average accuracy of 92.07% ± 5.72% and 82.22% ± 8.50%, for the ResNet-18 and Inflated 3D CNN algorithms, respectively. Contrastingly, LOEO provided a reduced average accuracy of 85.37% ± 14.38% and 77.79% ± 9.76%, for the ResNet-18 and Inflated 3D CNN algorithms, respectively.ConclusionThese pilot results indicate that the proposed algorithms can accurately distinguish between in-bed postures, on unseen participant data as well as unseen mattress environment data. The proposed algorithms can establish the basis of a decubitus ulcer prevention platform that can be applied to various sleeping environments. To the best of our knowledge, the impact of mattress stiffness has not been considered in previous studies regarding in-bed posture monitoring.

Parkinson’s disease detection based on features refinement through L1 regularized SVM and deep neural network

In previous studies, replicated and multiple types of speech data have been used for Parkinson’s disease (PD) detection. However, two main problems in these studies are lower PD detection accuracy and inappropriate validation methodologies leading to unreliable results. This study discusses the effects of inappropriate validation methodologies used in previous studies and highlights the use of appropriate alternative validation methods that would ensure generalization. To enhance PD detection accuracy, we propose a two-stage diagnostic system that refines the extracted set of features through L1\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$L_{1}$$\\end{document} regularized linear support vector machine and classifies the refined subset of features through a deep neural network. To rigorously evaluate the effectiveness of the proposed diagnostic system, experiments are performed on two different voice recording-based benchmark datasets. For both datasets, the proposed diagnostic system achieves 100% accuracy under leave-one-subject-out (LOSO) cross-validation (CV) and 97.5% accuracy under k-fold CV. The results show that the proposed system outperforms the existing methods regarding PD detection accuracy. The results suggest that the proposed diagnostic system is essential to improving non-invasive diagnostic decision support in PD.

Generalizing Upper Limb Force Modeling With Transfer Learning: A Multimodal Approach Using EMG and IMU for New Users and Conditions.

In the field of EMG-based force modeling, the ability to generalize models across individuals could play a significant role in its adoption across a range of applications, including assistive devices, robotic and rehabilitation devices. However, current studies have predominately focused on intra-subject modeling, largely neglecting the burden of end-user data acquisition. In this work, we propose the use of transfer learning (TL) to generalize force modeling to a new user by first establishing a baseline model trained using other users' data, and then adapting to the end-user using a small amount of new data (only 10% , 20% , and 40% of the new user data). Using a deep multimodal convolutional neural network, consisting of two CNN models, one with high-density (HD) EMG and one with motion data recorded by an Inertial Measurement Unit (IMU), our proposed TL technique significantly improved force modeling compared to leave-one-subject-out (LOSO) and even intra-subject scenarios. The TL approach increased the average R squared values of the force modeling task by 60.81%, 190.53%, and 199.79% compared to the LOSO case, and by 13.4%, 36.88%, and 45.51% compared to the intra-subject case for isotonic, isokinetic and dynamic conditions, respectively. These results show that it is possible to adapt to a new user with minimal data while improving performance significantly compared to the intra-subject scenario. We also show that TL can be used to generalize on a new experimental condition for a new user.