Detection of emotional states plays a prominent role in affective computing, decision-making, and healthcare. Physiological signals are an ideal target for continuous and objective assessment of emotional states. Electrodermal activity (EDA) is considered one of the most effective and widely used markers of sympathetic activation. However, emotional state assessment using EDA is challenging. We hypothesize that this is caused by the impossibility of current EDA analysis tools to capture the non-linear variations of the signal. Selecting an appropriate non-linear signal processing method will allow more effective emotion recognition using EDA signals. In this study, we have compared the performance of four non-linear signal processing methods— Improved symbolic aggregate approximation (isaxEDA), Complexity analysis (comEDA), Topological analysis (topEDA), and Network theory-based analysis (netEDA)—for classifying emotional states using EDA signals. For this, EDA signals were obtained from the publicly available Continuously Annotated Signals of Emotion (CASE) dataset. Features are extracted from each of the non-linear signal processing methods, encompassing aspects related to the dynamics, complexity, structure, and network properties of EDA. Subsequently, the extracted features from each method were employed as input for four machine learning classifiers: Naïve Bayes, Random Forest, k-Nearest Neighbors, and Support Vector Machine, which were validated using leave-one-subject-out cross-validation. The results demonstrate that isaxEDA achieved the highest performance, with an F1-score of 65% using an SVM classifier. The study supports the suitability of non-linear EDA signal processing to improve emotion recognition, which could be used to detect mental health conditions such as anxiety and depression.

The continuous detection of emotional states has many applications in mental health, marketing, human-computer interaction, and assistive robotics. Electrodermal activity (EDA), a signal modulated by sympathetic nervous system activity, provides continuous insight into emotional states. However, EDA possesses intricate nonstationary and nonlinear characteristics, making the extraction of emotion-relevant information challenging. We propose a novel graph signal processing (GSP) approach to model EDA signals as graphical networks, termed EDA-graph. The GSP leverages graph theory concepts to capture complex relationships in time-series data. To test the usefulness of EDA-graphs to detect emotions, we processed EDA recordings from the CASE emotion dataset using GSP by quantizing and linking values based on the Euclidean distance between the nearest neighbors. From these EDA-graphs, we computed the features of graph analysis, including total load centrality (TLC), total harmonic centrality (THC), number of cliques (GNC), diameter, and graph radius, and compared those features with features obtained using traditional EDA processing techniques. EDA-graph features encompassing TLC, THC, GNC, diameter, and radius demonstrated significant differences (p<0.05) between five emotional states (Neutral, Amused, Bored, Relaxed, and Scared). Using machine learning models for classifying emotional states evaluated using leave-one-subject-out cross-validation, we achieved a five-class F1 score of up to 0.68.

Electroretinograms (ERGs) show differences between typically developing populations and those with a diagnosis of autism spectrum disorder (ASD) or attention deficit/hyperactivity disorder (ADHD). In a series of ERGs collected in ASD (n = 77), ADHD (n = 43), ASD + ADHD (n = 21), and control (n = 137) groups, this analysis explores the use of machine learning and feature selection techniques to improve the classification between these clinically defined groups. Standard time domain and signal analysis features were evaluated in different machine learning models. For ASD classification, a balanced accuracy (BA) of 0.87 was achieved for male participants. For ADHD, a BA of 0.84 was achieved for female participants. When a three-group model (ASD, ADHD, and control) the BA was lower, at 0.70, and fell further to 0.53 when all groups were included (ASD, ADHD, ASD + ADHD, and control). The findings support a role for the ERG in establishing a broad two-group classification of ASD or ADHD, but the model’s performance depends upon sex and is limited when multiple classes are included in machine learning modeling.

Objective: Smartwatches with photoplethysmographic (PPG) sensors are ideal for early atrial fibrillation (AF) detection through continuous monitoring. However, prior deep learning was limited either to controlled environments, to minimize motion artifacts, or to short duration data collection. Additionally, premature atrial/ventricular contractions (PAC/PVC), which often confound AF detection algorithms, remains understudied due to limited datasets. Current state-of-the-art methods achieve only 75% sensitivity for PAC/PVC class on minimally motion artifact corrupted PPG data, despite showing 97% AF detection accuracy. Methods: We addressed the above limitations using data from the recently completed NIH-funded Pulsewatch clinical trial which collected over two weeks of smartwatch PPG data from 106 subjects. Our computationally efficient 1D bi-directional Gated Recurrent Unit deep learning model incorporated multi-modal inputs (1D PPG, accelerometer, and heart rate data) to classify normal sinus rhythm, AF, and PAC/PVC. Results: Our model achieved an unprecedented 83% sensitivity for PAC/PVC detection while maintaining a high accuracy of 97.31% for AF detection, outperforming the best retrained state-of-the-art model by 20.81% and 2.55%, respectively. It was also 14 times more computationally efficient and 2.7 times faster. Testing on two external PPG datasets collected with a different smartwatch and a fingertip PPG sensor, our model demonstrated better generalizability with macro-averaged AUROC values of 96.22% and 94.17%, respectively. Conclusion: A light-weight multimodal input deep learning model can accurately distinguish PAC/PVC from AF, reducing false positive detection of AF. Significance: Accurate AF and PAC/PVC detection with minimal false positive detection can enhance clinical and public acceptance of smartwatch-based AF monitoring.

The field of emotion recognition from physiological signals is a growing area of research with significant implications for both mental health monitoring and human–computer interaction. This study introduces a novel approach to detecting emotional states based on fractal analysis of electrodermal activity (EDA) signals. We employed detrended fluctuation analysis (DFA), Hurst exponent estimation, and wavelet entropy calculation to extract fractal features from EDA signals obtained from the CASE dataset, which contains physiological recordings and continuous emotion annotations from 30 participants. The analysis revealed significant differences in fractal features across five emotional states (neutral, amused, bored, relaxed, and scared), particularly those derived from wavelet entropy. A cross-correlation analysis showed robust correlations between fractal features and both the arousal and valence dimensions of emotion, challenging the conventional view of EDA as a predominantly arousal-indicating measure. The application of machine learning for emotion classification using fractal features achieved a leave-one-subject-out accuracy of 84.3% and an F1 score of 0.802, surpassing the performance of previous methods on the same dataset. This study demonstrates the potential of fractal analysis in capturing the intricate, multi-scale dynamics of EDA signals for emotion recognition, opening new avenues for advancing emotion-aware systems and affective computing applications.

Normal pressure hydrocephalus (NPH) is a neurological disorder characterized by altered cerebrospinal fluid accumulation in the brain’s ventricles, leading to symptoms such as gait disturbance and cognitive impairment. Artificial intelligence (AI), including machine learning (ML) and deep learning (DL), shows promise in diagnosing NPH using medical images. In this systematic review, we examined 21 papers on the use of AI in detecting NPH. The studies primarily focused on differentiating NPH from other neurodegenerative disorders, such as Parkinson’s disease and Alzheimer’s disease. We found that traditional ML methods like Support Vector Machines, Random Forest, and Logistic Regression were commonly used, while DL methods, particularly Deep Convolutional Neural Networks, were also widely employed. The accuracy of these approaches varied, ranging from 70% to 95% in differentiating NPH from other conditions. Feature selection techniques were used to identify relevant parameters for diagnosis. MRI scans were more frequently used than CT scans, but both modalities showed promise. Evaluation metrics like Dice similarity coefficients and ROC-AUC were the most typical metrics of model performance. Challenges in implementing AI in clinical practice were identified, and the authors suggested that a hybrid deep-traditional ML framework could enhance NPH diagnosis. Further research is needed to maximize the benefits of AI while addressing limitations.