The strength of anticipated distractors shapes EEG alpha and theta oscillations in a Working Memory task.

Working Memory (WM) requires maintenance of task-relevant information and suppression of task-irrelevant/distracting information. Alpha and theta oscillations have been extensively investigated in relation to WM. However, studies that examine both theta and alpha bands in relation to distractors, encompassing not only power modulation but also connectivity modulation, remain scarce. Here, we depicted, at the EEG-source level, the increase in power and connectivity in theta and alpha bands induced by strong relative to weak distractors during a visual Sternberg-like WM task involving the encoding of verbal items. During retention, a strong or weak distractor was presented, predictable in time and nature. Analysis focused on the encoding and retention phases before distractor presentation. Theta and alpha power were computed in cortical regions of interest, and connectivity networks estimated via spectral Granger causality and synthetized using in/out degree indices. The following modulations were observed for strong vs. weak distractors. In theta band during encoding, the power in frontal regions increased, together with frontal-to-frontal and bottom-up occipital-to-temporal-to-frontal connectivity; even during retention, bottom-up theta connectivity increased. In alpha band during retention, but not during encoding, the power in temporal-occipital regions increased, together with top-down frontal-to-occipital and temporal-to-occipital connectivity. From our results, we postulate a proactive cooperation between theta and alpha mechanisms: the first would mediate enhancement of target representation both during encoding and retention, and the second would mediate increased inhibition of sensory areas during retention only, to suppress the processing of imminent distractor without interfering with the processing of ongoing target stimulus during encoding.

SpeechBrain-MOABB: An open-source Python library for benchmarking deep neural networks applied to EEG signals.

Deep learning has revolutionized EEG decoding, showcasing its ability to outperform traditional machine learning models. However, unlike other fields, EEG decoding lacks comprehensive open-source libraries dedicated to neural networks. Existing tools (MOABB and braindecode) prevent the creation of robust and complete decoding pipelines, as they lack support for hyperparameter search across the entire pipeline, and are sensitive to fluctuations in results due to network random initialization. Furthermore, the absence of a standardized experimental protocol exacerbates the reproducibility crisis in the field. To address these limitations, we introduce SpeechBrain-MOABB, a novel open-source toolkit carefully designed to facilitate the development of a comprehensive EEG decoding pipeline based on deep learning. SpeechBrain-MOABB incorporates a complete experimental protocol that standardizes critical phases, such as hyperparameter search and model evaluation. It natively supports multi-step hyperparameter search for finding the optimal hyperparameters in a high-dimensional space defined by the entire pipeline, and multi-seed training and evaluation for obtaining performance estimates robust to the variability caused by random initialization. SpeechBrain-MOABB outperforms other libraries, including MOABB and braindecode, with accuracy improvements of 14.9% and 25.2% (on average), respectively. By enabling easy-to-use and easy-to-share decoding pipelines, our toolkit can be exploited by neuroscientists for decoding EEG with neural networks in a replicable and trustworthy way.

Convolutional neural networks reveal properties of reach-to-grasp encoding in posterior parietal cortex.

Deep neural networks (DNNs) are widely adopted to decode motor states from both non-invasively and invasively recorded neural signals, e.g., for realizing brain-computer interfaces. However, the neurophysiological interpretation of how DNNs make the decision based on the input neural activity is limitedly addressed, especially when applied to invasively recorded data. This reduces decoder reliability and transparency, and prevents the exploitation of decoders to better comprehend motor neural encoding. Here, we adopted an explainable artificial intelligence approach - based on a convolutional neural network and an explanation technique - to reveal spatial and temporal neural properties of reach-to-grasping from single-neuron recordings of the posterior parietal area V6A. The network was able to accurately decode 5 different grip types, and the explanation technique automatically identified the cells and temporal samples that most influenced the network prediction. Grip encoding in V6A neurons already started at movement preparation, peaking during movement execution. A difference was found within V6A: dorsal V6A neurons progressively encoded more for increasingly advanced grips, while ventral V6A neurons for increasingly rudimentary grips, with both subareas following a linear trend between the amount of grip encoding and the level of grip skills. By revealing the elements of the neural activity most relevant for each grip with no a priori assumptions, our approach supports and advances current knowledge about reach-to-grasp encoding in V6A, and it may represent a general tool able to investigate neural correlates of motor or cognitive tasks (e.g., attention and memory tasks) from single-neuron recordings.

Decoding movement kinematics from EEG using an interpretable convolutional neural network.

Continuous decoding of hand kinematics has been recently explored for the intuitive control of electroencephalography (EEG)-based Brain-Computer Interfaces (BCIs). Deep neural networks (DNNs) are emerging as powerful decoders, for their ability to automatically learn features from lightly pre-processed signals. However, DNNs for kinematics decoding lack in the interpretability of the learned features and are only used to realize within-subject decoders without testing other training approaches potentially beneficial for reducing calibration time, such as transfer learning. Here, we aim to overcome these limitations by using an interpretable convolutional neural network (ICNN) to decode 2-D hand kinematics (position and velocity) from EEG in a pursuit tracking task performed by 13 participants. The ICNN is trained using both within-subject and cross-subject strategies, and also testing the feasibility of transferring the knowledge learned on other subjects on a new one. Moreover, the network eases the interpretation of learned spectral and spatial EEG features. Our ICNN outperformed most of the other state-of-the-art decoders, showing the best trade-off between performance, size, and training time. Furthermore, transfer learning improved kinematics prediction in the low data regime. The network attributed the highest relevance for decoding to the delta-band across all subjects, and to higher frequencies (alpha, beta, low-gamma) for a cluster of them; contralateral central and parieto-occipital sites were the most relevant, reflecting the involvement of sensorimotor, visual and visuo-motor processing. The approach improved the quality of kinematics prediction from the EEG, at the same time allowing interpretation of the most relevant spectral and spatial features.

Deep learning applied to EEG source-data reveals both ventral and dorsal visual stream involvement in holistic processing of social stimuli.

Perception of social stimuli (faces and bodies) relies on "holistic" (i.e., global) mechanisms, as supported by picture-plane inversion: perceiving inverted faces/bodies is harder than perceiving their upright counterpart. Albeit neuroimaging evidence suggested involvement of face-specific brain areas in holistic processing, their spatiotemporal dynamics and selectivity for social stimuli is still debated. Here, we investigate the spatiotemporal dynamics of holistic processing for faces, bodies and houses (adopted as control non-social category), by applying deep learning to high-density electroencephalographic signals (EEG) at source-level. Convolutional neural networks were trained to classify cortical EEG responses to stimulus orientation (upright/inverted), separately for each stimulus type (faces, bodies, houses), resulting to perform well above chance for faces and bodies, and close to chance for houses. By explaining network decision, the 150-200 ms time interval and few visual ventral-stream regions were identified as mostly relevant for discriminating face and body orientation (lateral occipital cortex, and for face only, precuneus cortex, fusiform and lingual gyri), together with two additional dorsal-stream areas (superior and inferior parietal cortices). Overall, the proposed approach is sensitive in detecting cortical activity underlying perceptual phenomena, and by maximally exploiting discriminant information contained in data, may reveal spatiotemporal features previously undisclosed, stimulating novel investigations.

Motor decoding from the posterior parietal cortex using deep neural networks.

Objective.Motor decoding is crucial to translate the neural activity for brain-computer interfaces (BCIs) and provides information on how motor states are encoded in the brain. Deep neural networks (DNNs) are emerging as promising neural decoders. Nevertheless, it is still unclear how different DNNs perform in different motor decoding problems and scenarios, and which network could be a good candidate for invasive BCIs.Approach.Fully-connected, convolutional, and recurrent neural networks (FCNNs, CNNs, RNNs) were designed and applied to decode motor states from neurons recorded from V6A area in the posterior parietal cortex (PPC) of macaques. Three motor tasks were considered, involving reaching and reach-to-grasping (the latter under two illumination conditions). DNNs decoded nine reaching endpoints in 3D space or five grip types using a sliding window approach within the trial course. To evaluate decoders simulating a broad variety of scenarios, the performance was also analyzed while artificially reducing the number of recorded neurons and trials, and while performing transfer learning from one task to another. Finally, the accuracy time course was used to analyze V6A motor encoding.Main results.DNNs outperformed a classic Naïve Bayes classifier, and CNNs additionally outperformed XGBoost and Support Vector Machine classifiers across the motor decoding problems. CNNs resulted the top-performing DNNs when using less neurons and trials, and task-to-task transfer learning improved performance especially in the low data regime. Lastly, V6A neurons encoded reaching and reach-to-grasping properties even from action planning, with the encoding of grip properties occurring later, closer to movement execution, and appearing weaker in darkness.Significance.Results suggest that CNNs are effective candidates to realize neural decoders for invasive BCIs in humans from PPC recordings also reducing BCI calibration times (transfer learning), and that a CNN-based data-driven analysis may provide insights about the encoding properties and the functional roles of brain regions.

Modulations of Cortical Power and Connectivity in Alpha and Beta Bands during the Preparation of Reaching Movements.

Planning goal-directed movements towards different targets is at the basis of common daily activities (e.g., reaching), involving visual, visuomotor, and sensorimotor brain areas. Alpha (8-13 Hz) and beta (13-30 Hz) oscillations are modulated during movement preparation and are implicated in correct motor functioning. However, how brain regions activate and interact during reaching tasks and how brain rhythms are functionally involved in these interactions is still limitedly explored. Here, alpha and beta brain activity and connectivity during reaching preparation are investigated at EEG-source level, considering a network of task-related cortical areas. Sixty-channel EEG was recorded from 20 healthy participants during a delayed center-out reaching task and projected to the cortex to extract the activity of 8 cortical regions per hemisphere (2 occipital, 2 parietal, 3 peri-central, 1 frontal). Then, we analyzed event-related spectral perturbations and directed connectivity, computed via spectral Granger causality and summarized using graph theory centrality indices (in degree, out degree). Results suggest that alpha and beta oscillations are functionally involved in the preparation of reaching in different ways, with the former mediating the inhibition of the ipsilateral sensorimotor areas and disinhibition of visual areas, and the latter coordinating disinhibition of the contralateral sensorimotor and visuomotor areas.

Aquatic Therapy after Incomplete Spinal Cord Injury: Gait Initiation Analysis Using Inertial Sensors.

Populations with potential damage to somatosensory, vestibular, and visual systems or poor motor control are often studied during gait initiation. Aquatic activity has shown to benefit the functional capacity of incomplete spinal cord injury (iSCI) patients. The present study aimed to evaluate gait initiation in iSCI patients using an easy-to-use protocol employing four wearable inertial sensors. Temporal and acceleration-based anticipatory postural adjustment measures were computed and compared between dry-land and water immersion conditions in 10 iSCI patients. In the aquatic condition, an increased first step duration (median value of 1.44 s vs. 0.70 s in dry-land conditions) and decreased root mean squared accelerations for the upper trunk (0.39 m/s2 vs. 0.72 m/s2 in dry-land conditions) and lower trunk (0.41 m/s2 vs. 0.85 m/s2 in dry-land conditions) were found in the medio-lateral and antero-posterior direction, respectively. The estimation of these parameters, routinely during a therapy session, can provide important information regarding different control strategies adopted in different environments.

A Bayesian-optimized design for an interpretable convolutional neural network to decode and analyze the P300 response in autism.

Objective.P300 can be analyzed in autism spectrum disorder (ASD) to derive biomarkers and can be decoded in brain-computer interfaces to reinforce ASD impaired skills. Convolutional neural networks (CNNs) have been proposed for P300 decoding, outperforming traditional algorithms but they (a) do not investigate optimal designs in different training conditions; (b) lack in interpretability. To overcome these limitations, an interpretable CNN (ICNN), that we recently proposed for motor decoding, has been modified and adopted here, with its optimal design searched via Bayesian optimization.Approach.The ICNN provides a straightforward interpretation of spectral and spatial features learned to decode P300. The Bayesian-optimized (BO) ICNN design was investigated separately for different training strategies (within-subject, within-session, and cross-subject) and BO models were used for the subsequent analyses. Specifically, transfer learning (TL) potentialities were investigated by assessing how pretrained cross-subject BO models performed on a new subject vs. random-initialized models. Furthermore, within-subject BO-derived models were combined with an explanation technique (ICNN + ET) to analyze P300 spectral and spatial features.Main results.The ICNN resulted comparable or even outperformed existing CNNs, at the same time being lighter. BO ICNN designs differed depending on the training strategy, needing more capacity as the training set variability increased. Furthermore, TL provided higher performance than networks trained from scratch. The ICNN + ET analysis suggested the frequency range [2, 5.8] Hz as the most relevant, and spatial features showed a right-hemispheric parietal asymmetry. The ICNN + ET-derived features, but not ERP-derived features, resulted significantly and highly correlated to autism diagnostic observation schedule clinical scores.Significance.This study substantiates the idea that a CNN can be designed both accurate and interpretable for P300 decoding, with an optimized design depending on the training condition. The novel ICNN-based analysis tool was able to better capture ASD neural signatures than traditional event-related potential analysis, possibly paving the way for identifying novel biomarkers.

Decoding sensorimotor information from superior parietal lobule of macaque via Convolutional Neural Networks.

Despite the well-recognized role of the posterior parietal cortex (PPC) in processing sensory information to guide action, the differential encoding properties of this dynamic processing, as operated by different PPC brain areas, are scarcely known. Within the monkey's PPC, the superior parietal lobule hosts areas V6A, PEc, and PE included in the dorso-medial visual stream that is specialized in planning and guiding reaching movements. Here, a Convolutional Neural Network (CNN) approach is used to investigate how the information is processed in these areas. We trained two macaque monkeys to perform a delayed reaching task towards 9 positions (distributed on 3 different depth and direction levels) in the 3D peripersonal space. The activity of single cells was recorded from V6A, PEc, PE and fed to convolutional neural networks that were designed and trained to exploit the temporal structure of neuronal activation patterns, to decode the target positions reached by the monkey. Bayesian Optimization was used to define the main CNN hyper-parameters. In addition to discrete positions in space, we used the same network architecture to decode plausible reaching trajectories. We found that data from the most caudal V6A and PEc areas outperformed PE area in the spatial position decoding. In all areas, decoding accuracies started to increase at the time the target to reach was instructed to the monkey, and reached a plateau at movement onset. The results support a dynamic encoding of the different phases and properties of the reaching movement differentially distributed over a network of interconnected areas. This study highlights the usefulness of neurons' firing rate decoding via CNNs to improve our understanding of how sensorimotor information is encoded in PPC to perform reaching movements. The obtained results may have implications in the perspective of novel neuroprosthetic devices based on the decoding of these rich signals for faithfully carrying out patient's intentions.

Osteonecrosis in paediatric acute lymphoblastic leukaemia: Incidence, risk factors, radiological patterns and evolution in a single-centre cohort.

Osteonecrosis (ON) is a well-known sequela of paediatric acute lymphoblastic leukaemia (ALL) treatment. Incidence differs substantially among studies and the clinical significance of radiological findings is not fully established. We analysed 256 consecutive patients with ALL treated in our Institution between October 2010 and December 2016. Within the cohort, 41 developed ON, with a mean 5-year cumulative incidence of 18.5 (standard error, SE, 5.7)% overall. The mean (SE) 5-year cumulative incidence of ON was 12.7 (2.1)% after censoring upon stem cell transplantation (SCT) and/or relapse. Patients aged ≥10 years and patients allocated to the high-risk stratum had a 10-fold and fivefold higher risk of ON respectively. The risk of ON was more than double in relapsed patients, whereas no significant impact of gender, immunophenotype and SCT was demonstrated. Multiple lesions (median four joints involved per patient) were detected by magnetic resonance imaging in all but one patient, with the knee being the most affected joint. Lesions affecting convex joint surfaces experienced the worst evolution, whereas most lesions on diaphyses and concave surfaces remained radiologically stable or disappeared during follow-up. ON has a high prevalence in paediatric ALL, presenting with multiple lesions. Lesions involving convex surfaces were at higher risk of radiological deterioration.

A Lightweight Multi-Scale Convolutional Neural Network for P300 Decoding: Analysis of Training Strategies and Uncovering of Network Decision.

Convolutional neural networks (CNNs), which automatically learn features from raw data to approximate functions, are being increasingly applied to the end-to-end analysis of electroencephalographic (EEG) signals, especially for decoding brain states in brain-computer interfaces (BCIs). Nevertheless, CNNs introduce a large number of trainable parameters, may require long training times, and lack in interpretability of learned features. The aim of this study is to propose a CNN design for P300 decoding with emphasis on its lightweight design while guaranteeing high performance, on the effects of different training strategies, and on the use of post-hoc techniques to explain network decisions. The proposed design, named MS-EEGNet, learned temporal features in two different timescales (i.e., multi-scale, MS) in an efficient and optimized (in terms of trainable parameters) way, and was validated on three P300 datasets. The CNN was trained using different strategies (within-participant and within-session, within-participant and cross-session, leave-one-subject-out, transfer learning) and was compared with several state-of-the-art (SOA) algorithms. Furthermore, variants of the baseline MS-EEGNet were analyzed to evaluate the impact of different hyper-parameters on performance. Lastly, saliency maps were used to derive representations of the relevant spatio-temporal features that drove CNN decisions. MS-EEGNet was the lightest CNN compared with the tested SOA CNNs, despite its multiple timescales, and significantly outperformed the SOA algorithms. Post-hoc hyper-parameter analysis confirmed the benefits of the innovative aspects of MS-EEGNet. Furthermore, MS-EEGNet did benefit from transfer learning, especially using a low number of training examples, suggesting that the proposed approach could be used in BCIs to accurately decode the P300 event while reducing calibration times. Representations derived from the saliency maps matched the P300 spatio-temporal distribution, further validating the proposed decoding approach. This study, by specifically addressing the aspects of lightweight design, transfer learning, and interpretability, can contribute to advance the development of deep learning algorithms for P300-based BCIs.

Deep learning-based EEG analysis: investigating P3 ERP components.

The neural processing of incoming stimuli can be analysed from the electroencephalogram (EEG) through event-related potentials (ERPs). The P3 component is largely investigated as it represents an important psychophysiological marker of psychiatric disorders. This is composed by several subcomponents, such as P3a and P3b, reflecting distinct but interrelated sensory and cognitive processes of incoming stimuli. Due to the low EEG signal-to-noise-ratio, ERPs emerge only after an averaging procedure across trials and subjects. Thus, this canonical ERP analysis lacks in the ability to highlight EEG neural signatures at the level of single-subject and single-trial. In this study, a deep learning-based workflow is investigated to enhance EEG neural signatures related to P3 subcomponents already at single-subject and at single-trial level. This was based on the combination of a convolutional neural network (CNN) with an explanation technique (ET). The CNN was trained using two different strategies to produce saliency representations enhancing signatures shared across subjects or more specific for each subject and trial. Cross-subject saliency representations matched the signatures already emerging from ERPs, i.e., P3a and P3b-related activity within 350-400 ms (frontal sites) and 400-650 ms (parietal sites) post-stimulus, validating the CNN+ET respect to canonical ERP analysis. Single-subject and single-trial saliency representations enhanced P3 signatures already at the single-trial scale, while EEG-derived representations at single-subject and single-trial level provided no or only mildly evident signatures. Empowering the analysis of P3 modulations at single-subject and at single-trial level, CNN+ET could be useful to provide insights about neural processes linking sensory stimulation, cognition and behaviour.

A global benchmark of algorithms for segmenting the left atrium from late gadolinium-enhanced cardiac magnetic resonance imaging.

Segmentation of medical images, particularly late gadolinium-enhanced magnetic resonance imaging (LGE-MRI) used for visualizing diseased atrial structures, is a crucial first step for ablation treatment of atrial fibrillation. However, direct segmentation of LGE-MRIs is challenging due to the varying intensities caused by contrast agents. Since most clinical studies have relied on manual, labor-intensive approaches, automatic methods are of high interest, particularly optimized machine learning approaches. To address this, we organized the 2018 Left Atrium Segmentation Challenge using 154 3D LGE-MRIs, currently the world's largest atrial LGE-MRI dataset, and associated labels of the left atrium segmented by three medical experts, ultimately attracting the participation of 27 international teams. In this paper, extensive analysis of the submitted algorithms using technical and biological metrics was performed by undergoing subgroup analysis and conducting hyper-parameter analysis, offering an overall picture of the major design choices of convolutional neural networks (CNNs) and practical considerations for achieving state-of-the-art left atrium segmentation. Results show that the top method achieved a Dice score of 93.2% and a mean surface to surface distance of 0.7 mm, significantly outperforming prior state-of-the-art. Particularly, our analysis demonstrated that double sequentially used CNNs, in which a first CNN is used for automatic region-of-interest localization and a subsequent CNN is used for refined regional segmentation, achieved superior results than traditional methods and machine learning approaches containing single CNNs. This large-scale benchmarking study makes a significant step towards much-improved segmentation methods for atrial LGE-MRIs, and will serve as an important benchmark for evaluating and comparing the future works in the field. Furthermore, the findings from this study can potentially be extended to other imaging datasets and modalities, having an impact on the wider medical imaging community.

BCIAUT-P300: A Multi-Session and Multi-Subject Benchmark Dataset on Autism for P300-Based Brain-Computer-Interfaces.

There is a lack of multi-session P300 datasets for Brain-Computer Interfaces (BCI). Publicly available datasets are usually limited by small number of participants with few BCI sessions. In this sense, the lack of large, comprehensive datasets with various individuals and multiple sessions has limited advances in the development of more effective data processing and analysis methods for BCI systems. This is particularly evident to explore the feasibility of deep learning methods that require large datasets. Here we present the BCIAUT-P300 dataset, containing 15 autism spectrum disorder individuals undergoing 7 sessions of P300-based BCI joint-attention training, for a total of 105 sessions. The dataset was used for the 2019 IFMBE Scientific Challenge organized during MEDICON 2019 where, in two phases, teams from all over the world tried to achieve the best possible object-detection accuracy based on the P300 signals. This paper presents the characteristics of the dataset and the approaches followed by the 9 finalist teams during the competition. The winner obtained an average accuracy of 92.3% with a convolutional neural network based on EEGNet. The dataset is now publicly released and stands as a benchmark for future P300-based BCI algorithms based on multiple session data.

A fully automated left atrium segmentation approach from late gadolinium enhanced magnetic resonance imaging based on a convolutional neural network.

BACKGROUND: Several studies suggest that the evaluation of left atrial (LA) fibrosis is a relevant information for the assessment of the appropriate strategy in catheter ablation in atrial fibrillation (AF). Late gadolinium enhanced (LGE) cardiac magnetic resonance imaging (MRI) is a non-invasive technique, which might be employed for the non-invasive quantification of LA myocardial fibrotic tissue in patients with AF. Nowadays, the analysis of LGE MRI relies on manual tracing of LA boundaries and this procedure is time-consuming and prone to high inter-observer variability given the different degrees of observers' experience, LA wall thickness and data resolution. Therefore, an automated segmentation approach of the atrial cavity for the quantification of scar tissue would be highly desirable. METHODS: This study focuses on the design of a fully automated LGE MRI segmentation pipeline which includes a convolutional neural network (CNN) based on the successful architecture U-Net. The CNN was trained, validated and tested end-to-end with the data available from the Statistical Atlases and Computational Modelling of the Heart 2018 Atrial Segmentation Challenge (100 cardiac data). Two different approaches were tested: using both stacks of 2-D axial slices and using 3-D data (with the appropriate changes in the baseline architecture). In the latter approach, thanks to the 3-D convolution operator, all the information underlying 3-D data can be exploited. Once the training was completed using 80 cardiac data, a post-processing step was applied on 20 predicted segmentations belonging to the test set. RESULTS: By applying the 2-D and 3-D approaches, average Dice coefficient and mean Hausdorff distances were 0.896, 0.914, and 8.98 mm, 8.34 mm, respectively. Volumes of the anatomical LA meshes from the automated analysis were highly correlated with the volumes from ground truth [2-D: r=0.978, y=0.94x+0.07, bias=3.5 ml (5.6%), SD=5.3 mL (8.5%); 3-D: r=0.982, y=0.92x+2.9, bias=2.1 mL (3.5%), SD=5.2 mL (8.4%)]. CONCLUSIONS: These results suggest the proposed approach is feasible and provides accurate results. Despite the increase of the number of trainable parameters, the proposed 3-D CNN learns better features leading to higher performance, feasible for a real clinical application.

Effect of walking speed during gait in water of healthy elderly.

BACKGROUND: Walking in water (WW) is frequently used as an aquatic exercise in rehabilitation programs for the elderly. Understanding gait characteristics of WW is of primary importance to effectively design specific water-based rehabilitation programs. Moreover, as walking speed in water is reduced with a possible effect on gait parameters, the age- and environment-related changes during WW have to be investigated considering the effects of instantaneous walking speed. RESEARCH QUESTION: how do gait kinematic characteristics differ in healthy elderly between WW and on land walking condition (LW)? Do elderly show different walking patterns compared to young adults? Can these kinematic changes be accounted only by the different environment/age or are they also related to walking speed? METHODS: Nine healthy elderly participants (73.5 ±â€¯5.8 years) were acquired during walking in WW and LW at two different speeds. Kinematic parameters were assessed with waterproofed inertial magnetic sensors using a validated protocol. The influence of environment, age and walking speed on gait parameters was investigated with linear mixed models. RESULTS: Shorter stride distances and longer stride durations were observed in WW compared to LW. In the sagittal plane, hip and knee joint showed larger flexion in WW (>10deg over the whole stride and ∼28deg at foot strike, respectively). Furthermore, lower walking speeds and stride distances were observed in elderly compared to young adults. In the sagittal plane, a slightly more flexed hip joint and a less plantarflexed ankle joint (∼9 deg) were observed in the elderly. SIGNIFICANCE: The results showed the importance of assessing the walking speed during WW, as gait parameters can vary not only for the effect environment but also due to different walking speeds.

Interpretable and lightweight convolutional neural network for EEG decoding: Application to movement execution and imagination.

Convolutional neural networks (CNNs) are emerging as powerful tools for EEG decoding: these techniques, by automatically learning relevant features for class discrimination, improve EEG decoding performances without relying on handcrafted features. Nevertheless, the learned features are difficult to interpret and most of the existing CNNs introduce many trainable parameters. Here, we propose a lightweight and interpretable shallow CNN (Sinc-ShallowNet), by stacking a temporal sinc-convolutional layer (designed to learn band-pass filters, each having only the two cut-off frequencies as trainable parameters), a spatial depthwise convolutional layer (reducing channel connectivity and learning spatial filters tied to each band-pass filter), and a fully-connected layer finalizing the classification. This convolutional module limits the number of trainable parameters and allows direct interpretation of the learned spectral-spatial​ features via simple kernel visualizations. Furthermore, we designed a post-hoc gradient-based technique to enhance interpretation by identifying the more relevant and more class-specific features. Sinc-ShallowNet was evaluated on benchmark motor-execution and motor-imagery datasets and against different design choices and training strategies. Results show that (i) Sinc-ShallowNet outperformed a traditional machine learning algorithm and other CNNs for EEG decoding; (ii) The learned spectral-spatial features matched well-known EEG motor-related activity; (iii) The proposed architecture performed better with a larger number of temporal kernels still maintaining a good compromise between accuracy and parsimony, and with a trialwise rather than a cropped training strategy. In perspective, the proposed approach, with its interpretative capacity, can be exploited to investigate cognitive/motor aspects whose EEG correlates are yet scarcely known, potentially characterizing their relevant features.

The Use of IMMUs in a Water Environment: Instrument Validation and Application of 3D Multi-Body Kinematic Analysis in Medicine and Sport.

The aims of the present study were the instrumental validation of inertial-magnetic measurements units (IMMUs) in water, and the description of their use in clinical and sports aquatic applications applying customized 3D multi-body models. Firstly, several tests were performed to map the magnetic field in the swimming pool and to identify the best volume for experimental test acquisition with a mean dynamic orientation error lower than 5°. Successively, the gait and the swimming analyses were explored in terms of spatiotemporal and joint kinematics variables. The extraction of only spatiotemporal parameters highlighted several critical issues and the joint kinematic information has shown to be an added value for both rehabilitative and sport training purposes. Furthermore, 3D joint kinematics applied using the IMMUs provided similar quantitative information than that of more expensive and bulky systems but with a simpler and faster setup preparation, a lower time consuming processing phase, as well as the possibility to record and analyze a higher number of strides/strokes without limitations imposed by the cameras.

Gait Kinematic Analysis in Water Using Wearable Inertial Magnetic Sensors.

Walking is one of the fundamental motor tasks executed during aquatic therapy. Previous kinematics analyses conducted using waterproofed video cameras were limited to the sagittal plane and to only one or two consecutive steps. Furthermore, the set-up and post-processing are time-consuming and thus do not allow a prompt assessment of the correct execution of the movements during the aquatic session therapy. The aim of the present study was to estimate the 3D joint kinematics of the lower limbs and thorax-pelvis joints in sagittal and frontal planes during underwater walking using wearable inertial and magnetic sensors. Eleven healthy adults were measured during walking both in shallow water and in dry-land conditions. Eight wearable inertial and magnetic sensors were inserted in waterproofed boxes and fixed to the body segments by means of elastic modular bands. A validated protocol (Outwalk) was used. Gait cycles were automatically segmented and selected if relevant intraclass correlation coefficients values were higher than 0.75. A total of 704 gait cycles for the lower limb joints were normalized in time and averaged to obtain the mean cycle of each joint, among participants. The mean speed in water was 40% lower than that of the dry-land condition. Longer stride duration and shorter stride distance were found in the underwater walking. In the sagittal plane, the knee was more flexed (≈ 23°) and the ankle more dorsiflexed (≈ 9°) at heel strike, and the hip was more flexed at toe-off (≈ 13°) in water than on land. On the frontal plane in the underwater walking, smoother joint angle patterns were observed for thorax-pelvis and hip, and ankle was more inversed at toe-off (≈ 7°) and showed a more inversed mean value (≈ 7°). The results were mainly explained by the effect of the speed in the water as supported by the linear mixed models analysis performed. Thus, it seemed that the combination of speed and environment triggered modifications in the joint angles in underwater gait more than these two factors considered separately. The inertial and magnetic sensors, by means of fast set-up and data analysis, can supply an immediate gait analysis report to the therapist during the aquatic therapy session.