Adaptive data collection for intraindividual studies affected by adherence.

Recently, the use of mobile technologies in ecological momentary assessments (EMAs) and interventions has made it easier to collect data suitable for intraindividual variability studies in the medical field. Nevertheless, especially when self-reports are used during the data collection process, there are difficulties in balancing data quality and the burden placed on the subject. In this paper, we address this problem for a specific EMA setting that aims to submit a demanding task to subjects at high/low values of a self-reported variable. We adopt a dynamic approach inspired by control chart methods and design optimization techniques to obtain an EMA triggering mechanism for data collection that considers both the individual variability of the self-reported variable and of the adherence. We test the algorithm in both a simulation setting and with real, large-scale data from a tinnitus longitudinal study. A Wilcoxon signed rank test shows that the algorithm tends to have both a higher F1 score and utility than a random schedule and a rule-based algorithm with static thresholds, which are the current state-of-the-art approaches. In conclusion, the algorithm is proven effective in balancing data quality and the burden placed on the participants, especially in studies where data collection is impacted by adherence.

Cognitive Performance, Quality and Quantity of Movement Reflect Psychological Symptoms in Adolescents.

The presentation of unhealthy psychological symptoms are rising sharply in adolescents. Detrimental lifestyle behaviours are proposed as both possible causes and consequences. This study set out to compare selected measures of quality and quantity of movement between adolescents with and without unhealthy psychological symptoms. Using a cross sectional design, 96 participants completed the study from a whole year group of 166, age (13.36 ± 0.48) male 50.6% from a secondary school in Oxfordshire, England as a part of a larger study (EPIC) between January and April 2018. Measures were taken of quality and quantity of movement: reaction/movement time, gait pattern & physical activity, alongside psychological symptoms. Differences in movement behaviour in relation to psychological symptom and emotional problem presentation were determined using ANOVA. In the event of a significant result for the main factor of each parameter, a Bonferroni -corrected post hoc test was conducted to show the difference between categories in each group. Results for both unhealthy psychological symptoms and emotional problems were grouped into four categories ('Close to average', 'slightly raised', 'high' and 'very high'). Early adolescents with very high unhealthy psychological symptoms had 16.79% slower reaction times (p = 0.003, ηp2 = 0.170), 13.43% smaller walk ratio (p = 0.007, ηp2 = 0.152), 7.13% faster cadence (p = 0.005, ηp2 = 0.149), 6.95% less step time (p = 0.007, ηp2 = 0.153) and 1.4% less vigorous physical activity (p = 0.04, ηp2 = 0.102) than children with close to average psychological symptoms. Early adolescents with very high emotional problems had 12.25% slower reaction times (p = 0.05, ηp2 = 0.081), 10.61% smaller walk ratio (p = 0.02, ηp2 = 0.108), 6.03% faster cadence (p = 0.01, ηp2 = 0.134), 6.07% shorter step time (p = 0.007, ηp2 = 0.141) and 1.78% less vigorous physical activity (p = 0.009, ηp2 = 0.136) than children with close to average emotional problems. Different movement quality and quantity of was present in adolescents with unhealthy psychological symptoms and emotional problems. We propose movement may be used to both monitor symptoms, and as a novel therapeutic behavioural approach. Further studies are required to confirm our findings.

Multicomponent analysis using a confocal Raman microscope.

Measuring the concentration of multiple chemical components in a low-volume aqueous mixture by Raman spectroscopy has received significant interest in the literature. All of the contributions to date focus on the design of optical systems that facilitate the recording of spectra with high signal-to-noise ratio by collecting as many Raman scattered photons as possible. In this study, the confocal Raman microscope setup is investigated for multicomponent analysis. Partial least-squares regression is used to quantify physiologically relevant aqueous mixtures of glucose, lactic acid, and urea. The predicted error is 17.81 mg/dL for glucose, 10.6 mg/dL for lactic acid, and 7.6 mg/dL for urea, although this can be improved with increased acquisition times. A theoretical analysis of the method is proposed, which relates the numerical aperture and the magnification of the microscope objective, as well as the confocal pinhole size, to the performance of the technique.

Technology in Strength and Conditioning Tracking Lower-Limb Exercises With Wearable Sensors.

Strength and conditioning (S&C) coaches offer expert guidance to help those they work with achieve their personal fitness goals. However, because of cost and availability issues, individuals are often left training without expert supervision. Recent developments in inertial measurement units (IMUs) and mobile computing platforms have allowed for the possibility of unobtrusive motion tracking systems and the provision of real-time individualized feedback regarding exercise performance. These systems could enable S&C coaches to remotely monitor sessions and help gym users record workouts. One component of these IMU systems is the ability to identify the exercises completed. In this study, IMUs were positioned on the lumbar spine, thighs, and shanks on 82 healthy participants. Participants completed 10 repetitions of the squat, lunge, single-leg squat, deadlift, and tuck jump with acceptable form. Descriptive features were extracted from the IMU signals for each repetition of each exercise, and these were used to train an exercise classifier. The exercises were detected with 99% accuracy when using signals from all 5 IMUs, 99% when using signals from the thigh and lumbar IMUs and 98% with just a single IMU on the shank. These results indicate that a single IMU can accurately distinguish between 5 common multijoint exercises.

Technology in Strength and Conditioning: Assessing Bodyweight Squat Technique With Wearable Sensors.

O'Reilly, MA, Whelan, DF, Ward, TE, Delahunt, E, and Caulfield, BM. Technology in strength and conditioning: assessing bodyweight squat technique with wearable sensors. J Strength Cond Res 31(8): 2303-2312, 2017-Strength and conditioning (S&C) coaches offer expert guidance to help those they work with achieve their personal fitness goals. However, it is not always practical to operate under the direct supervision of an S&C coach and consequently individuals are often left training without expert oversight. Recent developments in inertial measurement units (IMUs) and mobile computing platforms have allowed for the possibility of unobtrusive motion tracking systems and the provision of real-time individualized feedback regarding exercise performance. These systems could enable S&C coaches to remotely monitor sessions and help individuals record their workout performance. One aspect of such technologies is the ability to assess exercise technique and detect common deviations from acceptable exercise form. In this study, we investigate this ability in the context of a bodyweight (BW) squat exercise. Inertial measurement units were positioned on the lumbar spine, thighs, and shanks of 77 healthy participants. Participants completed repetitions of BW squats with acceptable form and 5 common deviations from acceptable BW squatting technique. Descriptive features were extracted from the IMU signals for each BW squat repetition, and these were used to train a technique classifier. Acceptable or aberrant BW squat technique can be detected with 98% accuracy, 96% sensitivity, and 99% specificity when using features derived from all 5 IMUs. A single IMU system can also distinguish between acceptable and aberrant BW squat biomechanics with excellent accuracy, sensitivity, and specificity. Detecting exact deviations from acceptable BW squatting technique can be achieved with 80% accuracy using a 5 IMU system and 72% accuracy when using a single IMU positioned on the right shank. These results suggest that IMU-based systems can distinguish between acceptable and aberrant BW squat technique with excellent accuracy with a single IMU system. Identification of exact deviations is also possible but multi-IMU systems outperform single IMU systems.

An exploration of EEG features during recovery following stroke - implications for BCI-mediated neurorehabilitation therapy.

BACKGROUND: Brain-Computer Interfaces (BCI) can potentially be used to aid in the recovery of lost motor control in a limb following stroke. BCIs are typically used by subjects with no damage to the brain therefore relatively little is known about the technical requirements for the design of a rehabilitative BCI for stroke. METHODS: 32-channel electroencephalogram (EEG) was recorded during a finger-tapping task from 10 healthy subjects for one session and 5 stroke patients for two sessions approximately 6 months apart. An off-line BCI design based on Filter Bank Common Spatial Patterns (FBCSP) was implemented to test and compare the efficacy and accuracy of training a rehabilitative BCI with both stroke-affected and healthy data. RESULTS: Stroke-affected EEG datasets have lower 10-fold cross validation results than healthy EEG datasets. When training a BCI with healthy EEG, average classification accuracy of stroke-affected EEG is lower than the average for healthy EEG. Classification accuracy of the late session stroke EEG is improved by training the BCI on the corresponding early stroke EEG dataset. CONCLUSIONS: This exploratory study illustrates that stroke and the accompanying neuroplastic changes associated with the recovery process can cause significant inter-subject changes in the EEG features suitable for mapping as part of a neurofeedback therapy, even when individuals have scored largely similar with conventional behavioural measures. It appears such measures can mask this individual variability in cortical reorganization. Consequently we believe motor retraining BCI should initially be tailored to individual patients.

A dataset of text prompts, videos and video quality metrics from generative text-to-video AI models.

Evaluating the quality of videos which have been automatically generated from text-to-video (T2V) models is important if the models are to produce plausible outputs that convince a viewer of their authenticity. This paper presents a dataset of 201 text prompts used to automatically generate 1,005 videos using 5 very recent T2V models namely Tune-a-Video, VideoFusion, Text-To-Video Synthesis, Text2Video-Zero and Aphantasia. The prompts are divided into short, medium and longer lengths. We also include the results of some commonly used metrics used to automatically evaluate the quality of those generated videos. These include each video's naturalness, the text similarity between the original prompt and an automatically generated text caption for the video, and the inception score which measures how realistic is each generated video. Each of the 1,005 generated videos was manually rated by 24 different annotators for alignment between the videos and their original prompts, as well as for the perception and overall quality of the video. The data also includes the Mean Opinion Scores (MOS) for alignment between the generated videos and the original prompts. The dataset of T2V prompts, videos and assessments can be reused by those building or refining text-to-video generation models to compare the accuracy, quality and naturalness of their new models against existing ones.

The Potential of Electrospinning to Enable the Realization of Energy-Autonomous Wearable Sensing Systems.

The market for wearable electronic devices is experiencing significant growth and increasing potential for the future. Researchers worldwide are actively working to improve these devices, particularly in developing wearable electronics with balanced functionality and wearability for commercialization. Electrospinning, a technology that creates nano/microfiber-based membranes with high surface area, porosity, and favorable mechanical properties for human in vitro and in vivo applications using a broad range of materials, is proving to be a promising approach. Wearable electronic devices can use mechanical, thermal, evaporative and solar energy harvesting technologies to generate power for future energy needs, providing more options than traditional sources. This review offers a comprehensive analysis of how electrospinning technology can be used in energy-autonomous wearable wireless sensing systems. It provides an overview of the electrospinning technology, fundamental mechanisms, and applications in energy scavenging, human physiological signal sensing, energy storage, and antenna for data transmission. The review discusses combining wearable electronic technology and textile engineering to create superior wearable devices and increase future collaboration opportunities. Additionally, the challenges related to conducting appropriate testing for market-ready products using these devices are also discussed.

Semi-Supervised Mixture Learning for Graph Neural Networks With Neighbor Dependence.

A graph neural network (GNN) is a powerful architecture for semi-supervised learning (SSL). However, the data-driven mode of GNNs raises some challenging problems. In particular, these models suffer from the limitations of incomplete attribute learning, insufficient structure capture, and the inability to distinguish between node attribute and graph structure, especially on label-scarce or attribute-missing data. In this article, we propose a novel framework, called graph coneighbor neural network (GCoNN), for node classification. It is composed of two modules: GCoNN Γ and GCoNN Γ° . GCoNN Γ is trained to establish the fundamental prototype for attribute learning on labeled data, while GCoNN [Formula: see text] learns neighbor dependence on transductive data through pseudolabels generated by GCoNN Γ . Next, GCoNN Γ is retrained to improve integration of node attribute and neighbor structure through feedback from GCoNN [Formula: see text] . GCoNN tends to convergence iteratively using such an approach. From a theoretical perspective, we analyze this iteration process from a generalized expectation-maximization (GEM) framework perspective which optimizes an evidence lower bound (ELBO) by amortized variational inference. Empirical evidence demonstrates that the state-of-the-art performance of the proposed approach outperforms other methods. We also apply GCoNN to brain functional networks, the results of which reveal response features across the brain which are physiologically plausible with respect to known language and visual functions.

A P300-Based Brain-Computer Interface for Improving Attention.

A Brain-computer Interface (BCI) can be used as a neurofeedback training tool to improve cognitive performance. BCIs aim to improve the effectiveness and efficiency of the conventional neurofeedback methods by focusing on the self-regulation of individualized neuromarkers rather than generic ones in a graphically appealing training environment. In this work, for the first time, we have modified a widely used P300-based speller BCI and used it as an engaging neurofeedack training game to enhance P300. According to the user's performance the game becomes more difficult in an adaptive manner, requiring the generation of a larger and stronger P300 (i.e., in terms of total energy) in response to target stimuli. Since the P300 is generated naturally without conscious effort in response to a target trial, unlike many rhythm-based neurofeedback tools, the ability to control the proposed P300-based neurofeedback training is obtained after a short calibration without undergoing tedious trial and error sessions. The performance of the proposed neurofeedback training was evaluated over a short time scale (approximately 30 min training) using 28 young adult participants who were randomly assigned to either the experimental group or the control group. In summary, our results show that the proposed P300-based BCI neurofeedback training yielded a significant enhancement in the ERP components of the target trials (i.e., 150-550 ms after the onset of stimuli which includes P300) as well as attenuation in the corresponding ERP components of the non-target trials. In addition, more centro-parietal alpha suppression was observed in the experimental group during the neurofeedback training as well as a post-training spatial attention task. Interestingly, a significant improvement in the response time of a spatial attention task performed immediately after the neurofeedback training was observed in the experimental group. This paper, as a proof-of-concept study, suggests that the proposed neurofeedback training tool is a promising tool for improving attention particularly for those who are at risk of attention deficiency.

Brain-computer interface using a simplified functional near-infrared spectroscopy system.

A brain-computer interface (BCI) is a device that allows a user to communicate with external devices through thought processes alone. A novel signal acquisition tool for BCIs is near-infrared spectroscopy (NIRS), an optical technique to measure localized cortical brain activity. The benefits of using this non-invasive modality are safety, portability and accessibility. A number of commercial multi-channel NIRS system are available; however we have developed a straightforward custom-built system to investigate the functionality of a fNIRS-BCI system. This work describes the construction of the device, the principles of operation and the implementation of a fNIRS-BCI application, 'Mindswitch' that harnesses motor imagery for control. Analysis is performed online and feedback of performance is presented to the user. Mindswitch presents a basic 'on/off' switching option to the user, where selection of either state takes 1 min. Initial results show that fNIRS can support simple BCI functionality and shows much potential. Although performance may be currently inferior to many EEG systems, there is much scope for development particularly with more sophisticated signal processing and classification techniques. We hope that by presenting fNIRS as an accessible and affordable option, a new avenue of exploration will open within the BCI research community and stimulate further research in fNIRS-BCIs.

Feature-Free Activity Classification of Inertial Sensor Data With Machine Vision Techniques: Method, Development, and Evaluation.

BACKGROUND: Inertial sensors are one of the most commonly used sources of data for human activity recognition (HAR) and exercise detection (ED) tasks. The time series produced by these sensors are generally analyzed through numerical methods. Machine learning techniques such as random forests or support vector machines are popular in this field for classification efforts, but they need to be supported through the isolation of a potentially large number of additionally crafted features derived from the raw data. This feature preprocessing step can involve nontrivial digital signal processing (DSP) techniques. However, in many cases, the researchers interested in this type of activity recognition problems do not possess the necessary technical background for this feature-set development. OBJECTIVE: The study aimed to present a novel application of established machine vision methods to provide interested researchers with an easier entry path into the HAR and ED fields. This can be achieved by removing the need for deep DSP skills through the use of transfer learning. This can be done by using a pretrained convolutional neural network (CNN) developed for machine vision purposes for exercise classification effort. The new method should simply require researchers to generate plots of the signals that they would like to build classifiers with, store them as images, and then place them in folders according to their training label before retraining the network. METHODS: We applied a CNN, an established machine vision technique, to the task of ED. Tensorflow, a high-level framework for machine learning, was used to facilitate infrastructure needs. Simple time series plots generated directly from accelerometer and gyroscope signals are used to retrain an openly available neural network (Inception), originally developed for machine vision tasks. Data from 82 healthy volunteers, performing 5 different exercises while wearing a lumbar-worn inertial measurement unit (IMU), was collected. The ability of the proposed method to automatically classify the exercise being completed was assessed using this dataset. For comparative purposes, classification using the same dataset was also performed using the more conventional approach of feature-extraction and classification using random forest classifiers. RESULTS: With the collected dataset and the proposed method, the different exercises could be recognized with a 95.89% (3827/3991) accuracy, which is competitive with current state-of-the-art techniques in ED. CONCLUSIONS: The high level of accuracy attained with the proposed approach indicates that the waveform morphologies in the time-series plots for each of the exercises is sufficiently distinct among the participants to allow the use of machine vision approaches. The use of high-level machine learning frameworks, coupled with the novel use of machine vision techniques instead of complex manually crafted features, may facilitate access to research in the HAR field for individuals without extensive digital signal processing or machine learning backgrounds.

Classification of lunge biomechanics with multiple and individual inertial measurement units.

Lunges are a common, compound lower limb resistance exercise. If completed with aberrant technique, the increased stress on the joints used may increase risk of injury. This study sought to first investigate the ability of inertial measurement units (IMUs), when used in isolation and combination, to (a) classify acceptable and aberrant lunge technique (b) classify exact deviations in lunge technique. We then sought to investigate the most important features and establish the minimum number of top-ranked features and decision trees that are needed to maintain maximal system classification efficacy. Eighty volunteers performed the lunge with acceptable form and 11 deviations. Five IMUs positioned on the lumbar spine, thighs, and shanks recorded these movements. Time and frequency domain features were extracted from the IMU data and used to train and test a variety of classifiers. A single-IMU system achieved 83% accuracy, 62% sensitivity, and 90% specificity in binary classification and a five-IMU system achieved 90% accuracy, 80% sensitivity, and 92% specificity. A five-IMU set-up can also detect specific deviations with 70% accuracy. System efficiency was improved and classification quality was maintained when using only 20% of the top-ranked features for training and testing classifiers.

Classification of deadlift biomechanics with wearable inertial measurement units.

The deadlift is a compound full-body exercise that is fundamental in resistance training, rehabilitation programs and powerlifting competitions. Accurate quantification of deadlift biomechanics is important to reduce the risk of injury and ensure training and rehabilitation goals are achieved. This study sought to develop and evaluate deadlift exercise technique classification systems utilising Inertial Measurement Units (IMUs), recording at 51.2Hz, worn on the lumbar spine, both thighs and both shanks. It also sought to compare classification quality when these IMUs are worn in combination and in isolation. Two datasets of IMU deadlift data were collected. Eighty participants first completed deadlifts with acceptable technique and 5 distinct, deliberately induced deviations from acceptable form. Fifty-five members of this group also completed a fatiguing protocol (3-Repition Maximum test) to enable the collection of natural deadlift deviations. For both datasets, universal and personalised random-forests classifiers were developed and evaluated. Personalised classifiers outperformed universal classifiers in accuracy, sensitivity and specificity in the binary classification of acceptable or aberrant technique and in the multi-label classification of specific deadlift deviations. Whilst recent research has favoured universal classifiers due to the reduced overhead in setting them up for new system users, this work demonstrates that such techniques may not be appropriate for classifying deadlift technique due to the poor accuracy achieved. However, personalised classifiers perform very well in assessing deadlift technique, even when using data derived from a single lumbar-worn IMU to detect specific naturally occurring technique mistakes.

Technology in Rehabilitation: Comparing Personalised and Global Classification Methodologies in Evaluating the Squat Exercise with Wearable IMUs.

BACKGROUND: The barbell squat is a popularly used lower limb rehabilitation exercise. It is also an integral exercise in injury risk screening protocols. To date athlete/patient technique has been assessed using expensive laboratory equipment or subjective clinical judgement; both of which are not without shortcomings. Inertial measurement units (IMUs) may offer a low cost solution for the objective evaluation of athlete/patient technique. However, it is not yet known if global classification techniques are effective in identifying naturally occurring, minor deviations in barbell squat technique. OBJECTIVES: The aims of this study were to: (a) determine if in combination or in isolation, IMUs positioned on the lumbar spine, thigh and shank are capable of distinguishing between acceptable and aberrant barbell squat technique; (b) determine the capabilities of an IMU system at identifying specific natural deviations from acceptable barbell squat technique; and (c) compare a personalised (N=1) classifier to a global classifier in identifying the above. METHODS: Fifty-five healthy volunteers (37 males, 18 females, age = 24.21 +/- 5.25 years, height = 1.75 +/- 0.1 m, body mass = 75.09 +/- 13.56 kg) participated in the study. All participants performed a barbell squat 3-repetition maximum max strength test. IMUs were positioned on participants' lumbar spine, both shanks and both thighs; these were utilized to record tri-axial accelerometer, gyroscope and magnetometer data during all repetitions of the barbell squat exercise. Technique was assessed and labelled by a Chartered Physiotherapist using an evaluation framework. Features were extracted from the labelled IMU data. These features were used to train and evaluate both global and personalised random forests classifiers. RESULTS: Global classification techniques produced poor accuracy (AC), sensitivity (SE) and specificity (SP) scores in binary classification even with a 5 IMU set-up in both binary (AC: 64%, SE: 70%, SP: 28%) and multi-class classification (AC: 59%, SE: 24%, SP: 84%). However, utilising personalised classification techniques even with a single IMU positioned on the left thigh produced good binary classification scores (AC: 81%, SE: 81%, SP: 84%) and moderate-to-good multi-class scores (AC: 69%, SE: 70%, SP: 89%). CONCLUSIONS: There are a number of challenges in developing global classification exercise technique evaluation systems for rehabilitation exercises such as the barbell squat. Building large, balanced data sets to train such systems is difficult and time intensive. Minor, naturally occurring deviations may not be detected utilising global classification approaches. Personalised classification approaches allow for higher accuracy and greater system efficiency for end-users in detecting naturally occurring barbell squat technique deviations. Applying this approach also allows for a single-IMU set up to achieve similar accuracy to a multi-IMU setup, which reduces total system cost and maximises system usability.

Technology in Rehabilitation: Evaluating the Single Leg Squat Exercise with Wearable Inertial Measurement Units.

BACKGROUND: The single leg squat (SLS) is a common lower limb rehabilitation exercise. It is also frequently used as an evaluative exercise to screen for an increased risk of lower limb injury. To date athlete / patient SLS technique has been assessed using expensive laboratory equipment or subjective clinical judgement; both of which are not without shortcomings. Inertial measurement units (IMUs) may offer a low cost solution for the objective evaluation of athlete / patient SLS technique. OBJECTIVES: The aims of this study were to determine if in combination or in isolation IMUs positioned on the lumbar spine, thigh and shank are capable of: (a) distinguishing between acceptable and aberrant SLS technique; (b) identifying specific deviations from acceptable SLS technique. METHODS: Eighty-three healthy volunteers participated (60 males, 23 females, age: 24.68 + / - 4.91 years, height: 1.75 + / - 0.09 m, body mass: 76.01 + / - 13.29 kg). All participants performed 10 SLSs on their left leg. IMUs were positioned on participants' lumbar spine, left shank and left thigh. These were utilized to record tri-axial accelerometer, gyroscope and magnetometer data during all repetitions of the SLS. SLS technique was labelled by a Chartered Physiotherapist using an evaluation framework. Features were extracted from the labelled sensor data. These features were used to train and evaluate a variety of random-forests classifiers that assessed SLS technique. RESULTS: A three IMU system was moderately successful in detecting the overall quality of SLS performance (77 % accuracy, 77 % sensitivity and 78 % specificity). A single IMU worn on the shank can complete the same analysis with 76 % accuracy, 75 % sensitivity and 76 % specificity. Single sensors also produce competitive classification scores relative to multi-sensor systems in identifying specific deviations from acceptable SLS technique. CONCLUSIONS: A single IMU positioned on the shank can differentiate between acceptable and aberrant SLS technique with moderate levels of accuracy. It can also capably identify specific deviations from optimal SLS performance. IMUs may offer a low cost solution for the objective evaluation of SLS performance. Additionally, the classifiers described may provide useful input to an exercise biofeedback application.

Facilitating motor imagery-based brain-computer interface for stroke patients using passive movement.

Motor imagery-based brain-computer interface (MI-BCI) has been proposed as a rehabilitation tool to facilitate motor recovery in stroke. However, the calibration of a BCI system is a time-consuming and fatiguing process for stroke patients, which leaves reduced time for actual therapeutic interaction. Studies have shown that passive movement (PM) (i.e., the execution of a movement by an external agency without any voluntary motions) and motor imagery (MI) (i.e., the mental rehearsal of a movement without any activation of the muscles) induce similar EEG patterns over the motor cortex. Since performing PM is less fatiguing for the patients, this paper investigates the effectiveness of calibrating MI-BCIs from PM for stroke subjects in terms of classification accuracy. For this purpose, a new adaptive algorithm called filter bank data space adaptation (FB-DSA) is proposed. The FB-DSA algorithm linearly transforms the band-pass-filtered MI data such that the distribution difference between the MI and PM data is minimized. The effectiveness of the proposed algorithm is evaluated by an offline study on data collected from 16 healthy subjects and 6 stroke patients. The results show that the proposed FB-DSA algorithm significantly improved the classification accuracies of the PM and MI calibrated models (p < 0.05). According to the obtained classification accuracies, the PM calibrated models that were adapted using the proposed FB-DSA algorithm outperformed the MI calibrated models by an average of 2.3 and 4.5 % for the healthy and stroke subjects respectively. In addition, our results suggest that the disparity between MI and PM could be stronger in the stroke patients compared to the healthy subjects, and there would be thus an increased need to use the proposed FB-DSA algorithm in BCI-based stroke rehabilitation calibrated from PM.

Effects of feedback latency on P300-based brain-computer interface.

Feedback has been shown to affect performance when using a Brain-Computer Interface (BCI) based on sensorimotor rhythms. In contrast, little is known about the influence of feedback on P300-based BCIs. There is still an open question whether feedback affects the regulation of P300 and consequently the operation of P300-based BCIs. In this paper, for the first time, the influence of feedback on the P300-based BCI speller task is systematically assessed. For this purpose, 24 healthy participants performed the classic P300-based BCI speller task, while only half of them received feedback. Importantly, the number of flashes per letter was reduced on a regular basis in order to increase the frequency of providing feedback. Experimental results showed that feedback could significantly improve the P300-based BCI speller performance, if it was provided in short time intervals (e.g. in sequences as short as 4 to 6 flashes per row/column). Moreover, our offline analysis showed that providing feedback remarkably enhanced the relevant ERP patterns and attenuated the irrelevant ERP patterns, such that the discrimination between target and non-target EEG trials increased.

Slacking in the context of agent-based assessment in virtual rehabilitation systems.

Efforts are already underway to develop technology-derived solutions which automate aspects of conventional therapy. Ideally we would like to develop a human-like virtual therapist, in an attempt to enhance automated rehabilitation particularly in the home setting. One interesting skill of the experienced human therapist that we would like to model is the ability to recognize and manage behavior patterns known to decrease the effectiveness of rehabilitation. A particularly compelling example of such behavior is described in the context of robot-assisted therapy, where it has been demonstrated that "assist-as-needed" strategies may impact negatively on rehabilitation outcomes due to an intrinsic property of the human motor systems that encourages "slacking" as a form of energy optimization. In this work we endeavor to explore and extend this concept by giving it context in the standard therapist-patient interaction setting. We developed an apparatus which can measure and quantify grip strength and an agent based virtual therapist that can assess performance and offer simple natural language feedback in real time. We then conducted a series of experiments with healthy subjects in which the mapping between performance and feedback valence is altered. Our results demonstrate that subject performance is dependent on the feedback rules and that in particular, excessively positive feedback yields performance dynamics analogous to those observed in slacking studies. These preliminary results have implications for the design of virtual therapist systems.

Subject-to-subject adaptation to reduce calibration time in motor imagery-based brain-computer interface.

In order to enhance the usability of a motor imagery-based brain-computer interface (BCI), it is highly desirable to reduce the calibration time. Due to inter-subject variability, typically a new subject has to undergo a 20-30 minutes calibration session to collect sufficient data for training a BCI model based on his/her brain patterns. This paper proposes a new subject-to-subject adaptation algorithm to reliably reduce the calibration time of a new subject to only 3-4 minutes. To reduce the calibration time, unlike several past studies, the proposed algorithm does not require a large pool of historic sessions. In the proposed algorithm, using only a few trials from the new subject, first, the new subject's data is adapted to each available historic session separately. This is done by a linear transformation minimizing the distribution difference between the two groups of EEG data. Thereafter, among the available historic sessions, the one matched the most to the new subject's adapted data is selected as the calibration session. Consequently, the previously trained model based on the selected historic session is entirely used for the classification of the new subject's data after adaptation. The proposed algorithm is evaluated on a publicly available dataset with 9 subjects. For each subject, the calibration session is selected only from the calibration sessions of the eight other subjects. The experimental results showed that our proposed algorithm not only reduced the calibration time by 85%, but also performed on average only 1.7% less accurate than the subject-dependent calibration results.