Exploring the electrical robustness of conductive textile fasteners for wearable devices in different human motion conditions.

Conventional snap fasteners used in clothing are often used as electrical connectors in e-textile and wearable applications for signal transmission due to their wide availability and ease of use. Nonetheless, limited research exists on the validation of these fasteners, regarding the impact of contact-induced high-amplitude artefacts, especially under motion conditions. In this work, three types of fasteners were used as electromechanical connectors, establishing the interface between a regular sock and an acquisition device. The tested fasteners have different shapes and sizes, as well as have different mechanisms of attachment between the plug and receptacle counterparts. Experimental evaluation was performed under static conditions, slow walking, and rope jumping at a high cadence. The tests were also performed with a test mass of 140 g. Magnetic fasteners presented excellent electromechanical robustness under highly dynamic human movement with and without the additional mass. On the other hand, it was demonstrated that the Spring snap buttons (with a spring-based engaging mechanism) presented a sub-optimal performance under high motion and load conditions, followed by the Prong snap fasteners (without spring), which revealed a high susceptibility to artefacts. Overall, this work provides further evidence on the importance and reliability of clothing fasteners as electrical connectors in wearable systems.

Unsupervised analysis of whole transcriptome data from human pluripotent stem cells cardiac differentiation.

The main objective of the present work was to highlight differences and similarities in gene expression patterns between different pluripotent stem cell cardiac differentiation protocols, using a workflow based on unsupervised machine learning algorithms to analyse the transcriptome of cells cultured as a 2D monolayer or as 3D aggregates. This unsupervised approach effectively allowed to portray the transcriptomic changes that occurred throughout the differentiation processes, with a visual representation of the entire transcriptome. The results allowed to corroborate previously reported data and also to unveil new gene expression patterns. In particular, it was possible to identify a correlation between low cardiomyocyte differentiation efficiencies and the early expression of a set of non-mesodermal genes, which can be further explored as predictive markers of differentiation efficiency. The workflow here developed can also be applied to analyse other stem cell differentiation transcriptomic datasets, envisaging future clinical implementation of cellular therapies.

A Magnetic Field-Based Wearable Respiration Sensor for Real-Time Monitoring During Pulmonary Rehabilitation.

OBJECTIVE: This work explores Hall effect sensing paired with a permanent magnet, in the context of pulmonary rehabilitation exercise training. METHODS: Experimental evaluation was performed considering as reference the gold-standard of respiratory monitoring, an airflow transducer, and performance was compared to another wearable device with analogous usability - a piezoelectric sensor. A total of 16 healthy participants performed 15 activities, representative of pulmonary rehabilitation exercises, simultaneously using all devices. Evaluation was performed based on detection of flow reversal events and key respiratory parameters. RESULTS: Overall, the proposed sensor outperformed the piezoelectric sensor with a mean ratio, precision, and recall of 0.97, 0.97, and 0.95, respectively, against 0.98, 0.90, and 0.88. Evaluation regarding the respiratory parameters indicates an adequate accuracy when it comes to breath cycle, inspiration, and expiration times, with mean relative errors around 4% for breath cycle and 8% for inspiration/expiration times, despite some variability. Bland-Altman analysis indicates no systematic biases. CONCLUSION: Characterization of the proposed sensor shows adequate monitoring capabilities for exercises that do not rely heavily on torso mobility, but may present a limitation when it comes to activities such as side stretches. SIGNIFICANCE: This work provides a comprehensive characterization of a magnetic field-based respiration sensor, including a discussion on its robustness to different algorithm thresholds. It proves the viability of the sensor in a range of exercises, expanding the applicability of Hall effect sensors as a feasible wearable approach to real-time respiratory monitoring.

A real-world dataset of group emotion experiences based on physiological data.

Affective computing has experienced substantial advancements in recognizing emotions through image and facial expression analysis. However, the incorporation of physiological data remains constrained. Emotion recognition with physiological data shows promising results in controlled experiments but lacks generalization to real-world settings. To address this, we present G-REx, a dataset for real-world affective computing. We collected physiological data (photoplethysmography and electrodermal activity) using a wrist-worn device during long-duration movie sessions. Emotion annotations were retrospectively performed on segments with elevated physiological responses. The dataset includes over 31 movie sessions, totaling 380 h+ of data from 190+ subjects. The data were collected in a group setting, which can give further context to emotion recognition systems. Our setup aims to be easily replicable in any real-life scenario, facilitating the collection of large datasets for novel affective computing systems.

PreEpiSeizures: description and outcomes of physiological data acquisition using wearable devices during video-EEG monitoring in people with epilepsy.

The PreEpiSeizures project was created to better understand epilepsy and seizures through wearable technologies. The motivation was to capture physiological information related to epileptic seizures, besides Electroencephalography (EEG) during video-EEG monitorings. If other physiological signals have reliable information of epileptic seizures, unobtrusive wearable technology could be used to monitor epilepsy in daily life. The development of wearable solutions for epilepsy is limited by the nonexistence of datasets which could validate these solutions. Three different form factors were developed and deployed, and the signal quality was assessed for all acquired biosignals. The wearable data acquisition was performed during the video-EEG of patients with epilepsy. The results achieved so far include 59 patients from 2 hospitals totaling 2,721 h of wearable data and 348 seizures. Besides the wearable data, the Electrocardiogram of the hospital is also useable, totalling 5,838 h of hospital data. The quality ECG signals collected with the proposed wearable is equated with the hospital system, and all other biosignals also achieved state-of-the-art quality. During the data acquisition, 18 challenges were identified, and are presented alongside their possible solutions. Though this is an ongoing work, there were many lessons learned which could help to predict possible problems in wearable data collections and also contribute to the epilepsy community with new physiological information. This work contributes with original wearable data and results relevant to epilepsy research, and discusses relevant challenges that impact wearable health monitoring.

A novel approach to automatic seizure detection using computer vision and independent component analysis.

OBJECTIVE: Epilepsy is a neurological disease that affects ~50 million people worldwide, 30% of which have refractory epilepsy and recurring seizures, which may contribute to higher anxiety levels and poorer quality of life. Seizure detection may contribute to addressing some of the challenges associated with this condition, by providing information to health professionals regarding seizure frequency, type, and/or location in the brain, thereby improving diagnostic accuracy and medication adjustment, and alerting caregivers or emergency services of dangerous seizure episodes. The main focus of this work was the development of an accurate video-based seizure-detection method that ensured unobtrusiveness and privacy preservation, and provided novel approaches to reduce confounds and increase reliability. METHODS: The proposed approach is a video-based seizure-detection method based on optical flow, principal component analysis, independent component analysis, and machine learning classification. This method was tested on a set of 21 tonic-clonic seizure videos (5-30 min each, total of 4 h and 36 min of recordings) from 12 patients using leave-one-subject-out cross-validation. RESULTS: High accuracy levels were observed, namely a sensitivity and specificity of 99.06% ± 1.65% at the equal error rate and an average latency of 37.45 ± 1.31 s. When compared to annotations by health care professionals, the beginning and ending of seizures was detected with an average offset of 9.69 ± 0.97 s. SIGNIFICANCE: The video-based seizure-detection method described herein is highly accurate. Moreover, it is intrinsically privacy preserving, due to the use of optical flow motion quantification. In addition, given our novel independence-based approach, this method is robust to different lighting conditions, partial occlusions of the patient, and other movements in the video frame, thereby setting the base for accurate and unobtrusive seizure detection.

Morphological Autoencoders for Beat-by-Beat Atrial Fibrillation Detection Using Single-Lead ECG.

Engineered feature extraction can compromise the ability of Atrial Fibrillation (AFib) detection algorithms to deliver near real-time results. Autoencoders (AEs) can be used as an automatic feature extraction tool, tailoring the resulting features to a specific classification task. By coupling an encoder to a classifier, it is possible to reduce the dimension of the Electrocardiogram (ECG) heartbeat waveforms and classify them. In this work we show that morphological features extracted using a Sparse AE are sufficient to distinguish AFib from Normal Sinus Rhythm (NSR) beats. In addition to the morphological features, rhythm information was included in the model using a proposed short-term feature called Local Change of Successive Differences (LCSD). Using single-lead ECG recordings from two referenced public databases, and with features from the AE, the model was able to achieve an F1-score of 88.8%. These results show that morphological features appear to be a distinct and sufficient factor for detecting AFib in ECG recordings, especially when designed for patient-specific applications. This is an advantage over state-of-the-art algorithms that need longer acquisition times to extract engineered rhythm features, which also requires careful preprocessing steps. To the best of our knowledge, this is the first work that presents a near real-time morphological approach for AFib detection under naturalistic ECG acquisition with a mobile device.

Feasibility of Electrodermal Activity and Photoplethysmography Data Acquisition at the Foot Using a Sock Form Factor.

Wearable devices have been shown to play an important role in disease prevention and health management, through the multimodal acquisition of peripheral biosignals. However, many of these wearables are exposed, limiting their long-term acceptability by some user groups. To overcome this, a wearable smart sock integrating a PPG sensor and an EDA sensor with textile electrodes was developed. Using the smart sock, EDA and PPG measurements at the foot/ankle were performed in test populations of 19 and 15 subjects, respectively. Both measurements were validated by simultaneously recording the same signals with a standard device at the hand. For the EDA measurements, Pearson correlations of up to 0.95 were obtained for the SCL component, and a mean consensus of 69% for peaks detected in the two locations was obtained. As for the PPG measurements, after fine-tuning the automatic detection of systolic peaks, the index finger and ankle, accuracies of 99.46% and 87.85% were obtained, respectively. Moreover, an HR estimation error of 17.40±14.80 Beats-Per-Minute (BPM) was obtained. Overall, the results support the feasibility of this wearable form factor for unobtrusive EDA and PPG monitoring.

Pre-injury performance is most important for predicting the level of match participation after Achilles tendon ruptures in elite soccer players: a study using a machine learning classifier.

PURPOSE: Achilles tendon ruptures (ATR) are career-threatening injuries in elite soccer players due to the decreased sports performance they commonly inflict. This study presents an exploratory data analysis of match participation before and after ATRs and an evaluation of the performance of a machine learning (ML) model based on pre-injury features to predict whether a player will return to a previous level of match participation. METHODS: The website transfermarkt.com was mined, between January and March of 2021, for relevant entries regarding soccer players who suffered an ATR while playing in first or second leagues. The difference between average minutes played per match (MPM) 1 year before injury and between 1 and 2 years after the injury was used to identify patterns in match participation after injury. Clustering analysis was performed using k-means clustering. Predictions of post-injury match participation were made using the XGBoost classification algorithm. The performance of this model was evaluated using the area under the receiver operating characteristic curve (AUROC) and Brier score loss (BSL). RESULTS: Two hundred and nine players were included in the study. Data from 32,853 matches was analysed. Exploratory data analysis revealed that forwards, midfielders and defenders increased match participation during the first year after injury, with goalkeepers still improving at 2 years. Players were grouped into four clusters regarding the difference between MPMs 1 year before injury and between 1 and 2 years after the injury. These groups ranged between a severe decrease (n = 34; - 59 ± 13 MPM), moderate decrease (n = 75; - 25 ± 8 MPM), maintenance (n = 70; 0 ± 8 MPM), or increase (n = 30; 32 ± 13 MPM). Regarding the predictive model, the average AUROC after cross-validation was 0.81 ± 0.10, and the BSL was 0.12, with the most important features relating to pre-injury match participation. CONCLUSION: Most players take 1 year to reach peak match participation after an ATR. Good performance was attained using a ML classifier to predict the level of match participation following an ATR, with features related to pre-injury match participation displaying the highest importance. LEVEL OF EVIDENCE: I.

EpiBOX: An Automated Platform for Long-Term Biosignal Collection.

Biosignals represent a first-line source of information to understand the behavior and state of human biological systems, often used in machine learning problems. However, the development of healthcare-related algorithms that are both personalized and robust requires the collection of large volumes of data to capture representative instances of all possible states. While the rise of flexible biosignal acquisition solutions has enabled the expedition of data collection, they often require complicated frameworks or do not provide the customization required in some research contexts. As such, EpiBOX was developed as an open-source, standalone, and automated platform that enables the long-term acquisition of biosignals, passable to be operated by individuals with low technological proficiency. In particular, in this paper, we present an in-depth explanation of the framework, methods for the evaluation of its performance, and the corresponding findings regarding the perspective of the end-user. The impact of the network connection on data transfer latency was studied, demonstrating innocuous latency values for reasonable signal strengths and manageable latency values even when the connection was unstable. Moreover, performance profiling of the EpiBOX user interface (mobile application) indicates a suitable performance in all aspects, providing an encouraging outlook on adherence to the system. Finally, the experience of our research group is described as a use case, indicating a promising outlook regarding the use of the EpiBOX framework within similar contexts. As a byproduct of these features, our hope is that by empowering physicians, technicians, and monitored subjects to supervise the biosignal collection process, we enable researchers to scale biosignal collection.

Neuroorganoleptics: Organoleptic Testing Based on Psychophysiological Sensing.

There is an increasing interest, in consumer behaviour research related to food and beverage, in taking a step further from the traditional self-report questionnaires and organoleptic properties assessment. With the growing availability of psychophysiological data acquisition devices, and advancements in the study of the underlying signal sources seeking affective state assessment, the use of psychophysiological data analysis is a natural evolution in organoleptic testing. In this paper we propose a protocol for what can be defined as neuroorganoleptic analysis, a method that combines traditional approaches with psychophysiological data acquired during sensory testing. Our protocol was applied to a case study project named MobFood, where four samples of food were tested by a total of 83 participants, using preference and acceptance tasks, across three different sessions. Best practices and lessons learned regarding the laboratory setting and the acquisition of psychophysiological data were derived from this case study, which are herein described. Preliminary results show that certain Heart Rate Variability (HRV) features have a strong correlation with the preferences self-reported by the participants.

Morphological autoencoders for apnea detection in respiratory gating radiotherapy.

BACKGROUND AND OBJECTIVE: Respiratory gating training is a common technique to increase patient proprioception, with the goal of (e.g.) minimizing the effects of organ motion during radiotherapy. In this work, we devise a system based on autoencoders for classification of regular, apnea and unconstrained breathing patterns (i.e. multiclass). METHODS: Our approach is based on morphological analysis of the respiratory signals, using an autoencoder trained on regular breathing. The correlation between the input and output of the autoencoder is used to train and test several classifiers in order to select the best. Our approach is evaluated in a novel real-world respiratory gating biofeedback training dataset and on the Apnea-ECG reference dataset. RESULTS: Accuracies of 95 ± 3.5% and 87 ± 6.6% were obtained for two different datasets, in the classification of breathing and apnea. These results suggest the viability of a generalised model to characterise the breathing patterns under study. CONCLUSIONS: Using autoencoders to learn respiratory gating training patterns allows a data-driven approach to feature extraction, by focusing only on the signal's morphology. The proposed system is prone to be used in real-time and could potentially be transferred to other domains.

Emotion Assessment Using Feature Fusion and Decision Fusion Classification Based on Physiological Data: Are We There Yet?

Emotion recognition based on physiological data classification has been a topic of increasingly growing interest for more than a decade. However, there is a lack of systematic analysis in literature regarding the selection of classifiers to use, sensor modalities, features and range of expected accuracy, just to name a few limitations. In this work, we evaluate emotion in terms of low/high arousal and valence classification through Supervised Learning (SL), Decision Fusion (DF) and Feature Fusion (FF) techniques using multimodal physiological data, namely, Electrocardiography (ECG), Electrodermal Activity (EDA), Respiration (RESP), or Blood Volume Pulse (BVP). The main contribution of our work is a systematic study across five public datasets commonly used in the Emotion Recognition (ER) state-of-the-art, namely: (1) Classification performance analysis of ER benchmarking datasets in the arousal/valence space; (2) Summarising the ranges of the classification accuracy reported across the existing literature; (3) Characterising the results for diverse classifiers, sensor modalities and feature set combinations for ER using accuracy and F1-score; (4) Exploration of an extended feature set for each modality; (5) Systematic analysis of multimodal classification in DF and FF approaches. The experimental results showed that FF is the most competitive technique in terms of classification accuracy and computational complexity. We obtain superior or comparable results to those reported in the state-of-the-art for the selected datasets.

ECG Biometrics Using Deep Learning and Relative Score Threshold Classification.

The field of biometrics is a pattern recognition problem, where the individual traits are coded, registered, and compared with other database records. Due to the difficulties in reproducing Electrocardiograms (ECG), their usage has been emerging in the biometric field for more secure applications. Inspired by the high performance shown by Deep Neural Networks (DNN) and to mitigate the intra-variability challenges displayed by the ECG of each individual, this work proposes two architectures to improve current results in both identification (finding the registered person from a sample) and authentication (prove that the person is whom it claims) processes: Temporal Convolutional Neural Network (TCNN) and Recurrent Neural Network (RNN). Each architecture produces a similarity score, based on the prediction error of the former and the logits given by the last, and fed to the same classifier, the Relative Score Threshold Classifier (RSTC).The robustness and applicability of these architectures were trained and tested on public databases used by literature in this context: Fantasia, MIT-BIH, and CYBHi databases. Results show that overall the TCNN outperforms the RNN achieving almost 100%, 96%, and 90% accuracy, respectively, for identification and 0.0%, 0.1%, and 2.2% equal error rate (EER) for authentication processes. When comparing to previous work, both architectures reached results beyond the state-of-the-art. Nevertheless, the improvement of these techniques, such as enriching training with extra varied data and transfer learning, may provide more robust systems with a reduced time required for validation.

Positive and Negative Evidence Accumulation Clustering for Sensor Fusion: An Application to Heartbeat Clustering.

In this work, a new clustering algorithm especially geared towards merging data arising from multiple sensors is presented. The algorithm, called PN-EAC, is based on the ensemble clustering paradigm and it introduces the novel concept of negative evidence. PN-EAC combines both positive evidence, to gather information about the elements that should be grouped together in the final partition, and negative evidence, which has information about the elements that should not be grouped together. The algorithm has been validated in the electrocardiographic domain for heartbeat clustering, extracting positive evidence from the heartbeat morphology and negative evidence from the distances between heartbeats. The best result obtained on the MIT-BIH Arrhythmia database yielded an error of 1.44%. In the St. Petersburg Institute of Cardiological Technics 12-Lead Arrhythmia Database database (INCARTDB), an error of 0.601% was obtained when using two electrocardiogram (ECG) leads. When increasing the number of leads to 4, 6, 8, 10 and 12, the algorithm obtains better results (statistically significant) than with the previous number of leads, reaching an error of 0.338%. To the best of our knowledge, this is the first clustering algorithm that is able to process simultaneously any number of ECG leads. Our results support the use of PN-EAC to combine different sources of information and the value of the negative evidence.

Benchmarking of the BITalino biomedical toolkit against an established gold standard.

The low-cost multimodal platform BITalino is being increasingly used for educational and research purposes. However, there is still a lack of well-structured work comparing data acquired by this toolkit against a reference device, using established experimental protocols. This work intends to fill the said gap by benchmarking the performance of BITalino against the BioPac MP35 Student Lab Pro device. This work followed a methodical experimental protocol to acquire data from the two devices simultaneously. Four physiological signals were acquired: electrocardiography, electromyography, electrodermal activity and electroencephalography. Root mean square error and coefficient of determination were computed to analyse differences between BITalino and BioPac. Electrodermal activity signals were very similar for the two devices, even without applying any major signal processing techniques. For electrocardiography, a simple morphological comparison also revealed high similarity between devices, and this similarity increased after a common segmentation procedure was followed. Regarding electromyography and electroencephalography data, the approach consisted of comparing features extracted using common post-processing methods. The differences between BITalino and BioPac were again small. Overall, the results presented here show a close similarity between data acquired by the BITalino and by the reference device. This is an important validation step for all researchers working with this multimodal platform.

Biosignal-Based Multimodal Emotion Recognition in a Valence-Arousal Affective Framework Applied to Immersive Video Visualization.

Many emotion recognition schemes have been proposed in the state-of-the-art. They generally differ in terms of the emotion elicitation methods, target emotional states to recognize, data sources or modalities, and classification techniques. In this work several biosignals are explored for emotion assessment during immersive video visualization, collecting multimodal data from Electrocardiography (ECG), Electrodermal Activity (EDA), Blood Volume Pulse (BVP) and Respiration sensors. Participants reported their emotional state of the day (baseline), and provided self-assessment of the emotion experienced in each video through the Self-Assessment Manikin (SAM), in the valence-arousal space. Multiple physiological and statistical features extracted from the biosignals were used as inputs to an emotion recognition workflow, targeting user-independent classification with two classes per dimension. Support Vector Machines (SVM) were used, as it is considered one of the most promising classifiers in the field. The proposed approach lead to accuracies of 69.13% for arousal and 67.75% for valence, which are encouraging for further research with a larger training dataset and population.

Experimental characterization and analysis of the BITalino platforms against a reference device.

Low-cost hardware platforms for biomedical engineering are becoming increasingly available, which empower the research community in the development of new projects in a wide range of areas related with physiological data acquisition. Building upon previous work by our group, this work compares the quality of the data acquired by means of two different versions of the multimodal physiological computing platform BITalino, with a device that can be considered a reference. We acquired data from 5 sensors, namely Accelerometry (ACC), Electrocardiography (ECG), Electroencephalography (EEG), Electrodermal Activity (EDA) and Electromyography (EMG). Experimental evaluation shows that ACC, ECG and EDA data are highly correlated with the reference in what concerns the raw waveforms. When compared by means of their commonly used features, EEG and EMG data are also quite similar across the different devices.

BIT: Biosignal Igniter Toolkit.

The study of biosignals has had a transforming role in multiple aspects of our society, which go well beyond the health sciences domains to which they were traditionally associated with. While biomedical engineering is a classical discipline where the topic is amply covered, today biosignals are a matter of interest for students, researchers and hobbyists in areas including computer science, informatics, electrical engineering, among others. Regardless of the context, the use of biosignals in experimental activities and practical projects is heavily bounded by the cost, and limited access to adequate support materials. In this paper we present an accessible, albeit versatile toolkit, composed of low-cost hardware and software, which was created to reinforce the engagement of different people in the field of biosignals. The hardware consists of a modular wireless biosignal acquisition system that can be used to support classroom activities, interface with other devices, or perform rapid prototyping of end-user applications. The software comprehends a set of programming APIs, a biosignal processing toolbox, and a framework for real time data acquisition and postprocessing.

Check your biosignals here: a new dataset for off-the-person ECG biometrics.

The Check Your Biosignals Here initiative (CYBHi) was developed as a way of creating a dataset and consistently repeatable acquisition framework, to further extend research in electrocardiographic (ECG) biometrics. In particular, our work targets the novel trend towards off-the-person data acquisition, which opens a broad new set of challenges and opportunities both for research and industry. While datasets with ECG signals collected using medical grade equipment at the chest can be easily found, for off-the-person ECG data the solution is generally for each team to collect their own corpus at considerable expense of resources. In this paper we describe the context, experimental considerations, methods, and preliminary findings of two public datasets created by our team, one for short-term and another for long-term assessment, with ECG data collected at the hand palms and fingers.