Understanding the influence of confounding factors in myoelectric control for discrete gesture recognition

Discrete myoelectric control-based gesture recognition has recently gained interest as a possible input modality for many emerging ubiquitous computing applications. Unlike the continuous control commonly employed in powered prostheses, discrete systems seek to recognize the dynamic sequences associated with gestures to generate event-based inputs. More akin to those used in general-purpose human-computer interaction, these could include, for example, a flick of the wrist to dismiss a phone call or a double tap of the index finger and thumb to silence an alarm. Moelectric control systems have been shown to achieve near-perfect classification accuracy, but in highly constrained offline settings. Real-world, online systems are subject to 'confounding factors' (i.e. factors that hinder the real-world robustness of myoelectric control that are not accounted for during typical offline analyses), which inevitably degrade system performance, limiting their practical use. Although these factors have been widely studied in continuous prosthesis control, there has been little exploration of their impacts on discrete myoelectric control systems for emerging applications and use cases. Correspondingly, this work examines, for the first time, three confounding factors and their effect on the robustness of discrete myoelectric control: (1)limb position variability, (2)cross-day use, and a newly identified confound faced by discrete systems (3)gesture elicitation speed. Results from four different discrete myoelectric control architectures: (1) Majority Vote LDA, (2) Dynamic Time Warping, (3) an LSTM network trained with Cross Entropy, and (4) an LSTM network trained with Contrastive Learning, show that classification accuracy is significantly degraded (p<0.05) as a result of each of these confounds. This work establishes that confounding factors are a critical barrier that must be addressed to enable the real-world adoption of discrete myoelectric control for robust and reliable gesture recognition.

Context-informed incremental learning improves both the performance and resilience of myoelectric control

Despite its rich history of success in controlling powered prostheses and emerging commercial interests in ubiquitous computing, myoelectric control continues to suffer from a lack of robustness. In particular, EMG-based systems often degrade over prolonged use resulting in tedious recalibration sessions, user frustration, and device abandonment. Unsupervised adaptation is one proposed solution that updates a model's parameters over time based on its own predictions during real-time use to maintain robustness without requiring additional user input or dedicated recalibration. However, these strategies can actually accelerate performance deterioration when they begin to classify (and thus adapt) incorrectly, defeating their own purpose. To overcome these limitations, we propose a novel adaptive learning strategy, Context-Informed Incremental Learning (CIIL), that leverages in situ context to better inform the prediction of pseudo-labels. In this work, we evaluate these CIIL strategies in an online target acquisition task for two use cases: (1) when there is a lack of training data and (2) when a drastic and enduring alteration in the input space has occurred. A total of 32 participants were evaluated across the two experiments. The results show that the CIIL strategies significantly outperform the current state-of-the-art unsupervised high-confidence adaptation and outperform models trained with the conventional screen-guided training approach, even after a 45-degree electrode shift (p < 0.05). Consequently, CIIL has substantial implications for the future of myoelectric control, potentially reducing the training burden while bolstering model robustness, and leading to improved real-time control.

Machine learning enabled detection of COVID-19 pneumonia using exhaled breath analysis: a proof-of-concept study

Detection of the severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) relies on real-time-reverse-transcriptase polymerase chain reaction (RT-PCR) on nasopharyngeal swabs. The false-negative rate of RT-PCR can be high when viral burden and infection is localized distally in the lower airways and lung parenchyma. An alternate safe, simple and accessible method for sampling the lower airways is needed to aid in the early and rapid diagnosis of COVID-19 pneumonia. In a prospective unblinded observational study, patients admitted with a positive RT-PCR and symptoms of SARS-CoV-2 infection were enrolled from three hospitals in Ontario, Canada. Healthy individuals or hospitalized patients with negative RT-PCR and without respiratory symptoms were enrolled into the control group. Breath samples were collected and analyzed by laser absorption spectroscopy (LAS) for volatile organic compounds (VOCs) and classified by machine learning (ML) approaches to identify unique LAS-spectra patterns (breathprints) for SARS-CoV-2. Of the 135 patients enrolled, 115 patients provided analyzable breath samples. Using LAS-breathprints to train ML classifier models resulted in an accuracy of 72.2%-81.7% in differentiating between SARS-CoV2 positive and negative groups. The performance was consistent across subgroups of different age, sex, body mass index, SARS-CoV-2 variants, time of disease onset and oxygen requirement. The overall performance was higher than compared to VOC-trained classifier model, which had an accuracy of 63%-74.7%. This study demonstrates that a ML-based breathprint model using LAS analysis of exhaled breath may be a valuable non-invasive method for studying the lower airways and detecting SARS-CoV-2 and other respiratory pathogens. The technology and the ML approach can be easily deployed in any setting with minimal training. This will greatly improve access and scalability to meet surge capacity; allow early and rapid detection to inform therapy; and offers great versatility in developing new classifier models quickly for future outbreaks.

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

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

Decision-Change Informed Rejection Improves Robustness in Pattern Recognition-Based Myoelectric Control

Post-processing techniques have been shown to improve the quality of the decision stream generated by classifiers used in pattern-recognition-based myoelectric control. However, these techniques have largely been tested individually and on well-behaved, stationary data, failing to fully evaluate their trade-offs between smoothing and latency during dynamic use. Correspondingly, in this work, we survey and compare 8 different post-processing and decision stream improvement schemes in the context of continuous and dynamic class transitions: majority vote, Bayesian fusion, onset locking, outlier detection, confidence-based rejection, confidence scaling, prior adjustment, and adaptive windowing. We then propose two new temporally aware post-processing schemes that use changes in the decision and confidence streams to better reject uncertain decisions. Our decision-change informed rejection (DCIR) approach outperforms existing schemes during both steady-state and transitions based on error rates and decision stream volatility whether using conventional or deep classifiers. These results suggest that added robustness can be gained by appropriately leveraging temporal context in myoelectric control.

Novel Wearable HD-EMG Sensor With Shift-Robust Gesture Recognition Using Deep Learning

In this work, we present a hardware-software solution to improve the robustness of hand gesture recognition to confounding factors in myoelectric control. The solution includes a novel, full-circumference, flexible, 64-channel high-density electromyography (HD-EMG) sensor called EMaGer. The stretchable, wearable sensor adapts to different forearm sizes while maintaining uniform electrode density around the limb. Leveraging this uniformity, we propose novel array barrel-shifting data augmentation (ABSDA) approach used with a convolutional neural network (CNN), and an anti-aliased CNN (AA-CNN), that provides shift invariance around the limb for improved classification robustness to electrode movement, forearm orientation, and inter-session variability. Signals are sampled from a 4×16 HD-EMG array of electrodes at a frequency of 1 kHz and 16-bit resolution. Using data from 12 non-amputated participants, the approach is tested in response to sensor rotation, forearm rotation, and inter-session scenarios. The proposed ABSDA-CNN method improves inter-session accuracy by 25.67% on average across users for 6 gesture classes compared to conventional CNN classification. A comparison with other devices shows that this benefit is enabled by the unique design of the EMaGer array. The AA-CNN yields improvements of up to 63.05% accuracy over non-augmented methods when tested with electrode displacements ranging from -45 ° to +45 ° around the limb. Overall, this article demonstrates the benefits of co-designing sensor systems, processing methods, and inference algorithms to leverage synergistic and interdependent properties to solve state-of-the-art problems.

Feature Learning Networks for Floor Sensor-based Gait Recognition

Deep learning (DL) has become a powerful tool in many image classification applications but often requires large training sets to achieve high accuracy. For applications where the available data are limited, this can become a severely limiting factor in model performance. To address this limitation, feature learning network approaches that integrate traditional feature extraction methods with DL frameworks have been proposed. In this study, the performances of traditional methods: discrete wavelet transform (DWT), discrete cosine transform (DCT), independent component analysis (ICA), and principal component analysis (PCA); and their corresponding feature networks based on a convolutional neural network (CNN) framework: ScatNet (wavelet scattering network), DCTNet, ICANet, and PCANet, were investigated for use in pressure-based footstep recognition when the limited sample size is available for person authentication. The results show that the feature learning networks (90.6% accuracy) achieved significantly better performance on average than the conventional feature extraction methods (79.7% accuracy) (p < 0.05). Among the different feature networks, PCANet provided the best verification performance, with an accuracy of 92.2%. Feature learning networks are simple and effective approaches that can be a promising solution for applications like floor-based gait recognition in a security access scenario (such as workspace environment and border control) when small amounts of data are available for training models to differentiate between a larger group of users.

Feature Selection for Continuous within- and Cross-User EEG-Based Emotion Recognition

The monitoring of emotional state is important in the prevention and management of mental health problems and is increasingly being used to support affective computing. As such, researchers are exploring various modalities from which emotion can be inferred, such as through facial images or via electroencephalography (EEG) signals. Current research commonly investigates the performance of machine-learning-based emotion recognition systems by exposing users to stimuli that are assumed to elicit a single unchanging emotional response. Moreover, in order to demonstrate better results, many models are tested in evaluation frameworks that do not reflect realistic real-world implementations. Consequently, in this paper, we explore the design of EEG-based emotion recognition systems using longer, variable stimuli using the publicly available AMIGOS dataset. Feature engineering and selection results are evaluated across four different cross-validation frameworks, including versions of leave-one-movie-out (testing with a known user, but a previously unseen movie), leave-one-person-out (testing with a known movie, but a previously unseen person), and leave-one-person-and-movie-out (testing on both a new user and new movie). Results of feature selection lead to a 13% absolute improvement over comparable previously reported studies, and demonstrate the importance of evaluation framework on the design and performance of EEG-based emotion recognition systems.

A Better Framework for the Assessment of Performance and Stability of Co-Adaptive Myoelectric Systems

Co-adaptive myoelectric human-machine systems are a fairly recent, but promising, advancement in pattern recognition-based myoelectric control. Their performance and stability, however, are not fully understood due in part to a lack of proper assessment tools. Time-series based analyses are typically used despite the availability of techniques used in other fields that can robustly measure stability and performance. In this research, we leverage the success achieved by lower limb systems to improve the assessment framework of co-adaptive myoelectric systems by exploiting a key feature common between the two systems. The cyclical dynamics found in lower limbs are also apparent in co-adaptive myoelectric systems, allowing us to analyze their behavior using Poincaré maps. A 10-day experiment was designed and conducted to observe the effects of algorithm adaptation and myoelectric experience level on the performance of a co-adaptive myoelectric control system. Through Poincaré maps, we were able to identify learning effects, as well as oscillations and uncertainty in performance. Assessment of these seemingly random variations in performance led to the inference that co-adaptive systems can be chaotic. Modelling co-adaptive myoelectric systems as cyclical leads to the application of an improved framework to better assess and describe their dynamics and performance.

Using infrared spectroscopy to analyze breath of patients diagnosed with breast cancer

e13579

Background: Population-level screening programs aimed at early detection and treatment of breast cancer saves lives. Analyzing breath using infrared spectroscopy offers a highly sensitive, non-invasive, and cost-effective mechanism for identifying exhaled volatile organic chemicals, and it is hypothesized that it may identify differences in the “breathprint” of women with breast cancer relative to those without a breast cancer diagnosis. Methods: Alveolar breath samples (10 L) were collected using a Breathe BioMedical alveolar breath sampler onto Tenax TA sorbent tubes. Corresponding room air samples (10 L) were collected in the same manner. Absorption spectra of the samples at a desorb temperature of 75 °C were measured by infrared cavity ring-down spectroscopy (IR-CRDS), a highly sensitive method of measuring absorption coefficients due to trace volatile organic compounds (VOCs) present in exhaled breath. After subtracting room air absorption and ordering each measured spectrum by increasing wavelength, missing values were imputed using spline interpolation. The absorption spectra were then normalized using one of four techniques: min-max, vector, peak or standard normal variate normalization. The first derivatives of the normalized absorption coefficients (187 values in total) were then used as features for discriminating samples from subjects with breast cancer and controls. The most useful features were selected based on minimum redundancy and maximum relevance (mRMR) and were used to train a linear support vector machine (SVM) classifier. Performance of classification models was estimated based on two data splitting configurations, non-nested leave-one-out cross-validation (LOOCV) and nested LOOCV. These approaches provide upper and lower bounds of expected model performance. Classification performance was used for tuning the number of features included in each model. Results: The analysis of this study is based on the spectra obtained from 70 participants (38 breast cancer and 32 controls), collected at the Saint John Regional Hospital in New Brunswick, Canada. Table below shows the non-nested and nested performance characteristics of classifiers with the best performing normalization technique. The number of features given for the nested model is not an integer as it indicates an average across the cross-validation folds. Conclusions: These results suggest that the classification of alveolar breath using IR-CRDS is a promising technique for the detection of breast cancer. Performance of classification models. AUC is the area under the receiver operator characteristics curve.[Table: see text]

Infrared cavity ring-down spectroscopy for detecting non-small cell lung cancer in exhaled breath

Early diagnosis of lung cancer greatly improves the likelihood of survival and remission, but limitations in existing technologies like low-dose computed tomography have prevented the implementation of widespread screening programs. Breath-based solutions that seek disease biomarkers in exhaled volatile organic compound (VOC) profiles show promise as affordable, accessible and non-invasive alternatives to traditional imaging. In this pilot work, we present a lung cancer detection framework using cavity ring-down spectroscopy (CRDS), an effective and practical laser absorption spectroscopy technique that has the ability to advance breath screening into clinical reality. The main aims of this work were to 1) test the utility of infrared CRDS breath profiles for discriminating non-small cell lung cancer (NSCLC) patients from controls, 2) compare models with VOCs as predictors to those with patterns from the CRDS spectra (breathprints) as predictors, and 3) present a robust approach for identifying relevant disease biomarkers. First, based on a proposed learning curve technique that estimated the limits of a model's performance at multiple sample sizes (10-158), the CRDS-based models developed in this work were found to achieve classification performance comparable or superior to like mass spectroscopy and sensor-based systems. Second, using 158 collected samples (62 NSCLC subjects and 96 controls), the accuracy range for the VOC-based model was 65.19%-85.44% (51.61%-66.13% sensitivity and 73.96%-97.92% specificity), depending on the employed cross-validation technique. The model based on breathprint predictors generally performed better, with accuracy ranging from 71.52%-86.08% (58.06%-82.26% sensitivity and 80.21%-88.54% specificity). Lastly, using a protocol based on consensus feature selection, three VOCs (isopropanol, dimethyl sulfide, and butyric acid) and two breathprint features (from a local binary pattern transformation of the spectra) were identified as possible NSCLC biomarkers. This research demonstrates the potential of infrared CRDS breath profiles and the developed early-stage classification techniques for NSCLC biomarker detection and screening.

Compression of EMG Signals Using Deep Convolutional Autoencoders

Efficient storage and transmission of electromyogram (EMG) data are important for emerging applications such as telemedicine and big data, as a vital tool for further advancement of the field. However, due to limitations in internet speed and hardware resources, transmission and storage of EMG data are challenging. As a solution, this work proposes a new method for EMG data compression using deep convolutional autoencoders (CAE). Eight-channel EMG data from 10 subjects, and high density EMG data from 18 subjects, were investigated for compression. The CAE architecture was designed to extract an abstract data representation that is heavily compressed, but from which the salient information for classification can be effectively reconstructed. The proposed method attained efficient compression; for CR=1600, the average PRDN (percentage RMS difference normalized) was 31.5% and the wrist motions classification accuracy (CA) reduced roughly 5%. The CAE substantially outperformed the state-of-the-art high-efficiency video coding and a well known wavelet-thresholding compression technique. Moreover, by reducing the bit-resolution of the CAEs compressed data from 24 bits to 6 bits, an additional 4-fold compression was achieved without significant degradation of the reconstruction performance. Furthermore, the CAEs inter subject performance was promising; e.g. for CR=1600, the PRDN for the inter-subject case was only 2.6% less than that of the within-subject performance. The powerful EMG compression performance with remarkable reconstruction results reflects the CAEs potential as an automatic end to end approach with the ability to learn the complete encoding and decoding process. Furthermore, the excellent inter-subject performance demonstrates the generalizability and usability of the proposed approach.

Recurrent Convolutional Neural Networks as an Approach to Position-Aware Myoelectric Prosthesis Control

OBJECTIVE

Persons with normal arm function can perform complex wrist and hand movements over a wide range of limb positions. However, for those with transradial amputation who use myoelectric prostheses, control across multiple limb positions can be challenging, frustrating, and can increase the likelihood of device abandonment. In response, the goal of this research was to investigate recurrent convolutional neural network (RCNN)-based position-aware myoelectric prosthesis control strategies.

METHODS

Surface electromyographic (EMG) and inertial measurement unit (IMU) signals, obtained from 16 non-disabled participants wearing two Myo armbands, served as inputs to RCNN classification and regression models. Such models predicted movements (wrist flexion/extension and forearm pronation/supination), based on a multi-limb-position training routine. RCNN classifiers and RCNN regressors were compared to linear discriminant analysis (LDA) classifiers and support vector regression (SVR) regressors, respectively. Outcomes were examined to determine whether RCNN-based control strategies could yield accurate movement predictions, while using the fewest number of available Myo armband data streams.

RESULTS

An RCNN classifier (trained with forearm EMG data, and forearm and upper arm IMU data) predicted movements with 99.00% accuracy (versus the LDAs 97.67%). An RCNN regressor (trained with forearm EMG and IMU data) predicted movements with R2 values of 84.93% for wrist flexion/extension and 84.97% for forearm pronation/supination (versus the SVRs 77.26% and 60.73%, respectively). The control strategies that employed these models required fewer than all available data streams.

CONCLUSION

RCNN-based control strategies offer novel means of mitigating limb position challenges.

SIGNIFICANCE

This research furthers the development of improved position-aware myoelectric prosthesis control.

Data Pre-Processing of Infrared Spectral Breathprints for Lung Cancer Detection

Though breath analysis shows promise as a noninvasive and cost-effective approach to lung cancer screening, biomarkers in exhaled breath samples can be overwhelmed by irrelevant internal and environmental volatile organic compounds (VOCs). These extraneous VOCs can obscure the disease signature in a spectral breathprint, hindering the performance of pattern recognition models. In this work, pre-processing pipelines consisting of missing value replacement, detrending, and normalization techniques were evaluated to reduce these effects and enhance the features of interest in infrared cavity ring-down spectra. The best performing pipeline consisted of moving average detrending, linear interpolation for missing values, and vector normalization. This model achieved an average accuracy of 73.04% across five types of classifiers, exhibiting an 8.36% improvement compared to a baseline model (p < 0.05). A linear support vector machine classifier yielded the best performance (79.75% accuracy, 67.74% sensitivity, 87.50% specificity). This work can serve to guide pre-processing in future lung cancer breath research and, more broadly, in infrared laser absorption spectroscopy in general.

Machine Learning Model Validation for Early Stage Studies with Small Sample Sizes

In early stage biomedical studies, small datasets are common due to the high cost and difficulty of sample collection with human subjects. This complicates the validation of machine learning models, which are best suited for large datasets. In this work, we examined feature selection techniques, validation frameworks, and learning curve fitting for small simulated datasets with known underlying discriminability, with the aim of identifying a protocol for estimating and interpreting early stage model performance and for planning future studies. Of a variety of examined validation configurations, a nested cross-validation framework provided the most accurate reflection of the selected features' discriminability, but the relevant features were often not properly identified during the feature selection stage for datasets with small sample sizes. Ultimately, we recommend that: (1) filter-based feature selection methods should be used to minimize overfitting to noise-based features, (2) statistical exploration should be conducted on datasets as a whole to estimate the level of discriminability and the feasibility of the classification problems, and (3) learning curves should be employed using nested cross-validation performance estimates for forecasting accuracy at larger sample sizes and estimating the required number of samples to converge towards best performance. This work should serve as a guideline for researchers incorporating machine learning in small-scale pilot studies.

Verification-Based Design of a Robust EMG Wake Word

Surface electromyography (sEMG) signals are now commonly used in continuous myoelectric control of prostheses. More recently, researchers have considered EMG-based gesture recognition systems for human computer interaction research. These systems instead focus on recognizing discrete gestures (like a finger snap). The majority of works, however, have focused on improving multi-class performance, with little consideration for false activations from "other" classes. Consequently, they lack the robustness needed for real-world applications which generally require a single motion class such as a mouse click or a wake word. Furthermore, many works have borrowed the windowed classification schemes from continuous control, and thus fail to leverage the temporal structure of the gesture. In this paper, we propose a verification-based approach to creating a robust EMG wake word using one-class classifiers (Support Vector Data Description, One Class-Support Vector Machine, Dynamic Time Warping (DTW) & Hidden Markov Models). The area under the ROC curve (AUC) is used as a feature optimization objective as it provides a better representation of the verification performance. Equal error rate (EER) and AUC are then used as evaluation metrics. The results are computed using both window-based and temporal classifiers on a dataset consisting of five different gestures, with a best EER of 0.04 and AUC of 0.98, recorded using a DTW scheme. These results demonstrate a design framework that may benefit the development of more robust solutions for EMG-based wake words or input commands for a variety of interactive applications.

Nanoparticle surface-enhanced Raman spectroscopy as a noninvasive, label-free tool to monitor hematological malignancy

Aim: Monitoring minimal residual disease remains a challenge to the effective medical management of hematological malignancies; yet surface-enhanced Raman spectroscopy (SERS) has emerged as a potential clinical tool to do so. Materials & methods: We developed a cell-free, label-free SERS approach using gold nanoparticles (nanoSERS) to classify hematological malignancies referenced against two control cohorts: healthy and noncancer cardiovascular disease. A predictive model was built using machine-learning algorithms to incorporate disease burden scores for patients under standard treatment upon. Results: Linear- and quadratic-discriminant analysis distinguished three cohorts with 69.8 and 71.4% accuracies, respectively. A predictive nanoSERS model correlated (MSE = 1.6) with established clinical parameters. Conclusion: This study offers a proof-of-concept for the noninvasive monitoring of disease progression, highlighting the potential to incorporate nanoSERS into translational medicine.

A linearly extendible multi-artifact removal approach for improved upper extremity EEG-based motor imagery decoding

BACKGROUND AND OBJECTIVE

Non-invasive multichannel Electroencephalography (EEG) recordings provide an alternative source of neural information from which motor imagery (MI) patterns associated with limb movement intent can be decoded for use as control inputs for rehabilitation robots. The presence of multiple inherent dynamic artifacts in EEG signals, however, poses processing challenges for brain-computer interface (BCI) systems. A large proportion of the existing EEG signal preprocessing methods focus on isolating single artifact per time from an ensemble of EEG trials and require calibration and/or reference electrodes, resulting in increased complexity of their application to MI-EEG controlled rehabilitation devices in practical settings. Also, a few existing multi-artifacts removal methods though explored in other domains, they have rarely been investigated in the space of MI-EEG signals for multiple artifacts cancellation in a simultaneous manner.

APPROACH

Building on the premise of previous works, this study propose a semi-automatic EEG preprocessing method that combines Generalized Eigenvalue Decomposition driven by low-rank approximation and a Multi-channel Wiener Filter (GEVD-MWF) that employs a learning technique for simultaneous elimination of multiple artifacts from MI-EEG signals. The proposed method is applied to remove multiple artifacts from 64-channel EEG signals recorded from transhumeral amputees while they performed distinct classes of upper limb MI tasks before decoding their movement intent using a selection of features and machine learning algorithms.

MAIN RESULTS

Experimental results show that the proposed GEVD-MWF method yields significant improvements in MI decoding accuracies, in the range of 13.23%-41.21% compared to four existing popular artifact removal algorithms. Further investigation revealed that the GEVD-MWF approach enabled accuracies in the range of 90.44% - 99.67% using "single trial" EEG recordings, which could eliminate the need to record and process large ensembles of EEG trials as commonly required in some existing approaches. Additionally, using a variant of the sequential forward floating selection algorithm, a subset of 9 channels was used to obtain a decoding accuracy of 93.73%±1.58%.

SIGNIFICANCE

Given its improved performance, reduced data requirements, and feasibility with few channels, the proposed GEVD-MWF could potentially spur the development of effective real-time control strategies for multi-degree of freedom EEG-based miniaturized rehabilitation robotic interfaces.

Deep Cross-User Models Reduce the Training Burden in Myoelectric Control

The effort, focus, and time to collect data and train EMG pattern recognition systems is one of the largest barriers to their widespread adoption in commercial applications. In addition to multiple repetitions of motions, including exemplars of confounding factors during the training protocol has been shown to be critical for robust machine learning models. This added training burden is prohibitive for most regular use cases, so cross-user models have been proposed that could leverage inter-repetition variability supplied by other users. Existing cross-user models have not yet achieved performance levels sufficient for commercialization and require users to closely adhere to a training protocol that is impractical without expert guidance. In this work, we extend a previously reported adaptive domain adversarial neural network (ADANN) to a cross-subject framework that requires very little training data from the end-user. We compare its performance to single-repetition within-user training and the previous state-of-the-art cross-subject technique, canonical correlation analysis (CCA). ADANN significantly outperformed CCA for both intact-limb (86.8–96.2%) and amputee (64.1–84.2%) populations. Moreover, the ADANN adaptation computation time was substantially lower than the time otherwise devoted to conducting a full within-subject training protocol. This study shows that cross-user models, enabled by deep-learned adaptations, may be a viable option for improved generalized pattern recognition-based myoelectric control.

A Transferable Adaptive Domain Adversarial Neural Network for Virtual Reality Augmented EMG-Based Gesture Recognition

Within the field of electromyography-based (EMG) gesture recognition, disparities exist between the offline accuracy reported in the literature and the real-time usability of a classifier. This gap mainly stems from two factors: 1) The absence of a controller, making the data collected dissimilar to actual control. 2) The difficulty of including the four main dynamic factors (gesture intensity, limb position, electrode shift, and transient changes in the signal), as including their permutations drastically increases the amount of data to be recorded. Contrarily, online datasets are limited to the exact EMG-based controller used to record them, necessitating the recording of a new dataset for each control method or variant to be tested. Consequently, this paper proposes a new type of dataset to serve as an intermediate between offline and online datasets, by recording the data using a real-time experimental protocol. The protocol, performed in virtual reality, includes the four main dynamic factors and uses an EMG-independent controller to guide movements. This EMG-independent feedback ensures that the user is in-the-loop during recording, while enabling the resulting dynamic dataset to be used as an EMG-based benchmark. The dataset is comprised of 20 able-bodied participants completing three to four sessions over a period of 14 to 21 days. The ability of the dynamic dataset to serve as a benchmark is leveraged to evaluate the impact of different recalibration techniques for long-term (across-day) gesture recognition, including a novel algorithm, named TADANN. TADANN consistently and significantly ( [Formula: see text]) outperforms using fine-tuning as the recalibration technique.

Wearable EMG-Based Gesture Recognition Systems During Activities of Daily Living: An Exploratory Study

Recent advancements in wearable technologies have increased the potential for practical gesture recognition systems using electromyogram (EMG) signals. However, despite the high classification accuracies reported in many studies (> 90%), there is a gap between academic results and industrial success. This is in part because state-of-the-art EMG-based gesture recognition systems are commonly evaluated in highly-controlled laboratory environments, where users are assumed to be resting and performing one of a closed set of target gestures. In real world conditions, however, a variety of non-target gestures are performed during activities of daily living (ADLs), resulting in many false positive activations. In this study, the effect of ADLs on the performance of EMG-based gesture recognition using a wearable EMG device was investigated. EMG data for 14 hand and finger gestures, as well as continuous activity during uncontrolled ADLs (>10 hours in total) were collected and analyzed. Results showed that (1) the cluster separability of 14 different gestures during ADLs was 171 times worse than during rest; (2) the probability distributions of EMG features extracted from different ADLs were significantly different (p <; 0.05). (3) of the 14 target gestures, a right angle gesture (extension of the thumb and index finger) was least often inadvertently activated during ADLs. These results suggest that ADLs and other non-trained gestures must be taken into consideration when designing EMG-based gesture recognition systems.

Feasibility of Data-driven EMG Signal Generation using a Deep Generative Model

Despite recent advancements in the field of pattern recognition-based myoelectric control, the collection of a high quality training set remains a challenge limiting its adoption. This paper proposes a framework for a possible solution by augmenting short training protocols with subject-specific synthetic electromyography (EMG) data generated using a deep generative network, known as SinGAN. The aim of this work is to produce high quality synthetic data that could improve classification accuracy when combined with a limited training protocol. SinGAN was used to generate 1000 synthetic windows of EMG data from a single window of six different motions, and results were evaluated qualitatively, quantitatively, and in a classification task. Qualitative assessment of synthetic data was conducted via visual inspection of principal component analysis projections of real and synthetic feature space. Quantitative assessment of synthetic data revealed 11 of 32 synthetic features had similar location and scale to real features (using univariate two-sample Lepage tests); whereas multivariate distributions were found to be statistically different (p <0.05). Finally, the addition of these synthetic data to a brief training set of real data significantly improved classification accuracy in a cross-validation testing scheme by 5.4% (p <0.001).

Assessment of EMG Benchmark Data for Gesture Recognition Using the NinaPro Database

In recent years, many electromyography (EMG) benchmark databases have been made publicly available to the myoelectric control research community. Many small laboratories that lack the instrumentation, access, and experience needed to collect quality EMG data have used these benchmark datasets to explore and propose new signal processing and pattern recognition algorithms. It is widely accepted that noise contamination can affect the performance of myoelectric control systems, and so useful datasets should maintain good signal quality to ensure accurate results for proposed EMG-based gesture recognition systems. Despite the availability and adoption of benchmarks datasets, however, the quality of the EMG signals in these benchmarks has not yet been examined. In this study, the signal quality of the Non-Invasive Adaptive Prosthetics (NinaPro) dataset, the most widely known publicly available benchmark database to date, was comprehensively investigated with the goals of: 1) reporting the level of noise contamination in each NinaPro sub-dataset, 2) proposing signal quality criteria for assessing EMG datasets, 3) analyzing the effect of signal quality on classification performance, and 4) examining the quality of the data labels.

Comparison of Feedback Approaches to Improve Training in Partial Weight-Bearing

Assistive devices, including canes or crutches, are used in partial weight-bearing (PWB) to offload weight from limbs weakened by disease or injury, promote recovery, and prevent reinjury. While weight must be offloaded accurately to target loads prescribed by healthcare providers for maximum benefit, current training methods result in poor adherence. It is, however, currently unknown how best to provide feedback during training so that users can build an accurate internal model for PWB. In this work, we investigate seven feedback schemes using an instrumented cane, which vary the modality, timing, and the level of detail provided. We find that auditory schemes and a retrospective visual scheme outperform current clinical practices for PWB training. These findings provide results that can be applied directly to improve current clinical practices and provide valuable new insight into the design of feedback for training internal models in force-based motor control tasks. Clinically, this work presents a simple modification to clinical PWB training practices that can improve compliance by up to 75%, positively influencing rehabilitation outcomes and reducing the risk of complications.

Analysis of exhaled breath of lung cancer patients using infrared spectroscopy

e21037

Background: Currently low-dose computed tomography is used for lung cancer (LC) screening, but is limited by radiation exposure, cost, and a high false detection rate (1,2). An accurate, accessible and affordable screening technology is needed to improve detection of LC in high risk individuals. Methods: We conducted an unblinded, prospective cohort study on the effectiveness of a novel technology utilizing infrared absorption measurements via cavity ringdown spectroscopy (IR-CRDS) to differentiate the expired breath of treatment-naïve LC patients from controls without known cancer. Breath samples were taken from 100 LC patients and 98 control subjects but, only 62 non-small cell lung cancer (NSCLC) and 96 control samples were analyzed. Patients on treatment were eligible but, the protocol was amended to exclude these due to signal ambiguities. Samples were also excluded due to missing data, unclear histologic subtypes, or if they were classified as small cell LC samples to prevent obscuring the NSCLC signal. A piecewise cubic spline interpolation was used for the spectra with missing values (3). After first- and second-derivative spectra were computed to increase the information density, a one-dimensional local binary pattern extracted features from the spectra (4). Meaningful spectra-based features were selected using a minimum redundancy maximum relevance algorithm (5). Finally, a classification model was built using a support vector machine classifier (3). Results: The table below characterizes each cohort. The discriminant analysis differentiated between NSCLC and control cases with a cross validation accuracy of 86.1% (89.6% sensitivity and 80.7% specificity) using 20 selected spectra-based features. Conclusions: IR absorption measurements can be used to accurately discriminate between NSCLC and control participants. We continue to build our database to support more robust machine learning models. To our knowledge, this is the first time IR-CRDS has been used to differentiate between NSCLC and control cases. [Table: see text]

Fractal Analysis of Human Gait Variability via Stride Interval Time Series

Fractal analysis of stride interval time series is a useful tool in human gait research which could be used as a marker for gait adaptability, gait disorder, and fall risk among patients with movement disorders. This study is designed to systematically and comprehensively investigate two practical aspects of fractal analysis which significantly affect the outcome: the series length and the parameters used in the algorithm. The Hurst exponent, scaling exponent, and/or fractal dimension are computed from both simulated and experimental data using three fractal methods, namely detrended fluctuation analysis, box-counting dimension, and Higuchi's fractal dimension. The advantages and drawbacks of each method are discussed, in terms of biases and variability. The results demonstrate that a careful selection of fractal analysis methods and their parameters is required, which is dependent on the aim of study (either analyzing differences between experimental groups or estimating an accurate determination of fractal features). A set of guidelines for the selection of the fractal methods and the length of stride interval time series is provided, along with the optimal parameters for a robust implementation for each method.

Current Trends and Confounding Factors in Myoelectric Control: Limb Position and Contraction Intensity

This manuscript presents a hybrid study of a comprehensive review and a systematic(research) analysis. Myoelectric control is the cornerstone ofmany assistive technologies used in clinicalpractice, such as prosthetics and orthoses, and human-computer interaction, such as virtual reality control.Although the classification accuracy of such devices exceeds 90% in a controlled laboratory setting,myoelectric devices still face challenges in robustness to variability of daily living conditions.The intrinsic physiological mechanisms limiting practical implementations of myoelectric deviceswere explored: the limb position effect and the contraction intensity effect. The degradationof electromyography (EMG) pattern recognition in the presence of these factors was demonstratedon six datasets, where classification performance was 13% and 20% lower than the controlledsetting for the limb position and contraction intensity effect, respectively. The experimental designsof limb position and contraction intensity literature were surveyed. Current state-of-the-art trainingstrategies and robust algorithms for both effects were compiled and presented. Recommendationsfor future limb position effect studies include: the collection protocol providing exemplars of at least 6positions (four limb positions and three forearm orientations), three-dimensional space experimentaldesigns, transfer learning approaches, and multi-modal sensor configurations. Recommendationsfor future contraction intensity effect studies include: the collection of dynamic contractions, nonlinearcomplexity features, and proportional control.

Interpreting Deep Learning Features for Myoelectric Control: A Comparison With Handcrafted Features

Existing research on myoelectric control systems primarily focuses on extracting discriminative characteristics of the electromyographic (EMG) signal by designing handcrafted features. Recently, however, deep learning techniques have been applied to the challenging task of EMG-based gesture recognition. The adoption of these techniques slowly shifts the focus from feature engineering to feature learning. Nevertheless, the black-box nature of deep learning makes it hard to understand the type of information learned by the network and how it relates to handcrafted features. Additionally, due to the high variability in EMG recordings between participants, deep features tend to generalize poorly across subjects using standard training methods. Consequently, this work introduces a new multi-domain learning algorithm, named ADANN (Adaptive Domain Adversarial Neural Network), which significantly enhances (p = 0.00004) inter-subject classification accuracy by an average of 19.40% compared to standard training. Using ADANN-generated features, this work provides the first topological data analysis of EMG-based gesture recognition for the characterization of the information encoded within a deep network, using handcrafted features as landmarks. This analysis reveals that handcrafted features and the learned features (in the earlier layers) both try to discriminate between all gestures, but do not encode the same information to do so. In the later layers, the learned features are inclined to instead adopt a one-vs.-all strategy for a given class. Furthermore, by using convolutional network visualization techniques, it is revealed that learned features actually tend to ignore the most activated channel during contraction, which is in stark contrast with the prevalence of handcrafted features designed to capture amplitude information. Overall, this work paves the way for hybrid feature sets by providing a clear guideline of complementary information encoded within learned and handcrafted features.

A Deep Transfer Learning Approach to Reducing the Effect of Electrode Shift in EMG Pattern Recognition-Based Control

An important barrier to commercialization of pattern recognition myoelectric control of prostheses is the lack of robustness to confounding factors such as electrode shift, skin impedance variations, and learning effects. To overcome this challenge, a novel supervised adaptation approach based on transfer learning (TL) with convolutional neural networks (CNNs) is proposed which requires only a short training session (a few seconds for each class) to recalibrate the system. TL is proposed as a solution to the problem of insufficient calibration data due to short training times for both classification and regression-based control schemes. This approach was validated for electrode shift of roughly 2.5cm with 13 able-bodied subjects to estimate individual and combined wrist motions. With this method, the original CNN (trained before the shift) was fine-tuned with the calibration data from after shifting. The results show that the proposed technique outperforms training a CNN from scratch (random initialization of weights) or a support vector machine (SVM) using the minimal calibration data. Moreover, it demonstrates superior performance than previous LDA and QDA-based adaptation approaches. As the outcomes confirm, the proposed CNN TL method provides a practical solution for adaptation to external factors, improving the robustness of electromyogram (EMG) pattern recognition systems.

Fusion of Neuro-Signals and Dynamic Signatures for Person Authentication

Many biometric systems based on physiological traits such as ones facial characteristics, iris, and fingerprint have been developed for authentication purposes. Such security systems, however, commonly suffer from impersonation attacks such as obfuscation, abrasion, latent samples, and covert attack. More conventional behavioral methods, such as passwords and signatures, suffer from similar issues and can easily be spoofed. With growing levels of private data readily available across the internet, a more robust authentication system is needed for use in emerging technologies and mobile applications. In this paper, we present a novel multimodal biometric user authentication framework by combining the behavioral dynamic signature with the the physiological electroencephalograph (EEG) to restrict unauthorized access. EEG signals of 33 genuine users were collected while signing on their mobile phones. The recorded sequences were modeled using a bidirectional long short-term memory neural network (BLSTM-NN) based sequential classifier to accomplish person identification and verification. An accuracy of 98.78% was obtained for identification using decision fusion of dynamic signatures and EEG signals. The robustness of the framework was also tested against 1650 impersonation attempts made by 25 forged users by imitating the dynamic signatures of genuine users. Verification performance was measured using detection error tradeoff (DET) curves and half total error rate (HTER) security matrices using true positive rate (TPR) and false acceptance rate (FAR), resulting in 3.75% FAR and 1.87% HTER with 100% TPR for forgery attempts.

EMG Pattern Recognition for Persons with Cervical Spinal Cord Injury

Pattern recognition based myoelectric control has been widely explored in the field of prosthetics, but little work has extended to other patient groups. Individuals with neurological injuries such as spinal cord injury may also benefit from more intuitive control that may facilitate more interactive treatments or improved control of functional electrical stimulation (FES) systems or assistive technologies. This work presents a pilot study with 10 individuals with cervical spinal cord injury between A and C on the American Spinal Injury Association Impairment Scale. Subjects attempted to elicit 10 classes of forearm and hand movements while their electromyogram (EMG) was recorded using a cuff of eight electrodes. Various well-known EMG features were evaluated using a linear discriminant analysis classifier, yielding classification error rates as low as 4.3% ± 3.9 across the 10 classes. Reducing the number of classes to five, those required to control a commercial therapeutic FES device, further reduced the error rates to (2.2% ± 4.4). Results from this study provide evidence supporting continued exploration of EMG pattern recognition techniques for use by high-level spinal cord injured populations as a method of intuitive control over interactive FES systems or assistive devices.

Feature Extraction and Selection for Pain Recognition Using Peripheral Physiological Signals

In pattern recognition, the selection of appropriate features is paramount to both the performance and the robustness of the system. Over-reliance on machine learning-based feature selection methods can, therefore, be problematic; especially when conducted using small snapshots of data. The results of these methods, if adopted without proper interpretation, can lead to sub-optimal system design or worse, the abandonment of otherwise viable and important features. In this work, a deep exploration of pain-based emotion classification was conducted to better understand differences in the results of the related literature. In total, 155 different time domain and frequency domain features were explored, derived from electromyogram (EMG), skin conductance levels (SCL), and electrocardiogram (ECG) readings taken from the 85 subjects in response to heat-induced pain. To address the inconsistency in the optimal feature sets found in related works, an exhaustive and interpretable feature selection protocol was followed to obtain a generalizable feature set. Associations between features were then visualized using a topologically-informed chart, called Mapper, of this physiological feature space, including synthesis and comparison of results from previous literature. This topological feature chart was able to identify key sources of information that led to the formation of five main functional feature groups: signal amplitude and power, frequency information, nonlinear complexity, unique, and connecting. These functional groupings were used to extract further insight into observable autonomic responses to pain through a complementary statistical interaction analysis. From this chart, it was observed that EMG and SCL derived features could functionally replace those obtained from ECG. These insights motivate future work on novel sensing modalities, feature design, deep learning approaches, and dimensionality reduction techniques.

FMG Versus EMG: A Comparison of Usability for Real-Time Pattern Recognition Based Control

OBJECTIVE

Force myography (FMG), which measures the surface pressure profile exerted by contracting muscles, has been proposed as an alternative to electromyography (EMG) for human-machine interfaces. Although FMG pattern recognition-based control systems have yielded higher offline classification accuracy, comparatively few works have examined the usability of FMG for real-time control. In this work, we conduct a comprehensive comparison of EMG- and FMG-based schemes using both classification and regression controllers.

METHODS

A total of 20 participants performed a two-degree-of-freedom Fitts' Law-style virtual target acquisition task using both FMG- and EMG-based classification and regression control schemes. Performance was evaluated based on the standard Fitts' law testing metrics throughput, path efficiency, average speed, number of timeouts, overshoot, stopping distance, and simultaneity.

RESULTS

The FMG-based classification system significantly outperformed the EMG-based classification system in both throughput (0.902 ± 0.270) versus (0.751 ± 0.309), (ρ < 0.001) and path efficiency (87.2 ± 8.7) versus (83.2 ± 7.8), (ρ < 0.001). Similarly, FMG-based regression significantly outperformed EMG-based regression in throughput (0.871 ± 0.2) versus (0.69 ± 0.3), (ρ < 0.001) and path efficiency (64.8 ± 5.3) versus (58.8 ± 7.1), (ρ < 0.001).

CONCLUSIONS

The FMG-based schemes outperformed the EMG-based schemes regardless of which controller was used. This provides further evidence for FMG as a viable alternative to EMG for human-machine interfaces.

SIGNIFICANCE

This work describes a comprehensive evaluation of the online usability of FMG- and EMG-based control using both sequential classification and simultaneous regression control.

An Investigation of Temporally Inspired Time Domain Features for Electromyographic Pattern Recognition

This paper presents a novel set of temporally inspiredtime domain features for electromyographic (EMG) pattern recognition. The proposed methods employ simple time series measures derived from peak detection, and could better reflect EMG activity over time. Multiple EMG datasets consisting of 68 able-bodied and transradial amputee subjects performing a large variety of hand, wrist, fingers, and grasping movements are used to evaluate the performance of the proposed features and to design robust EMG feature sets. The results show that the average classification accuracy of two novel features, the mean prominence of local peaks and valleys, outperform several commonly used time domain features, autoregressive coefficients, histogram, and zero crossing, by 8{\%, 11{\%, and 17{\%, respectively. The proposed features are also shown to provide additional information as part of a robust feature set when compared to the common Hudgins' timedomain feature set, as selected by sequential forward selection and through empirical feature set design.

Real-time, simultaneous myoelectric control using a convolutional neural network

The evolution of deep learning techniques has been transformative as they have allowed complex mappings to be trained between control inputs and outputs without the need for feature engineering. In this work, a myoelectric control system based on convolutional neural networks (CNN) is proposed as a possible alternative to traditional approaches that rely on specifically designed features. This CNN-based system is validated using a real-time Fitts' law style target acquisition test requiring single and combined wrist motions. The performance of the proposed system is then compared to that of a standard support vector machine (SVM) based myoelectric system using a set of time-domain features. Despite the prevalence and demonstrated performance of these well-known features, no significant difference (p>0.05) was found between the two methods for any of the computed control metrics. This demonstrates the potential for automated learning approaches to extract complex and rich information from stochastic biological signals. This first evaluation of the usability of a CNN in a real-time myoelectric control environment provides a basis for further exploration.

A Multi-Sensor Matched Filter Approach to Robust Segmentation of Assisted Gait

Individuals with mobility impairments related to age, injury, or disease, often require the help of an assistive device (AD) such as a cane to ambulate, increase safety, and improve overall stability. Instrumenting these devices has been proposed as a non-invasive way to proactively monitor an individual’s reliance on the AD while also obtaining information about behaviors and changes in gait. A critical first step in the analysis of these data, however, is the accurate processing and segmentation of the sensor data to extract relevant gait information. In this paper, we present a highly accurate multi-sensor-based gait segmentation algorithm that is robust to a variety of walking conditions using an AD. A matched filtering approach based on loading information is used in conjunction with an angular rate reversal and peak detection technique, to identify important gait events. The algorithm is tested over a variety of terrains using a hybrid sensorized cane, capable of measuring loading, mobility, and stability information. The reliability and accuracy of the proposed multi-sensor matched filter (MSMF) algorithm is compared with variations of the commonly employed gyroscope peak detection (GPD) algorithm. Results of an experiment with a group of 30 healthy participants walking over various terrains demonstrated the ability of the proposed segmentation algorithm to reliably and accurately segment gait events.

Navigating features: a topologically informed chart of electromyographic features space

The success of biological signal pattern recognition depends crucially on the selection of relevant features. Across signal and imaging modalities, a large number of features have been proposed, leading to feature redundancy and the need for optimal feature set identification. A further complication is that, due to the inherent biological variability, even the same classification problem on different datasets can display variations in the respective optimal sets, casting doubts on the generalizability of relevant features. Here, we approach this problem by leveraging topological tools to create charts of features spaces. These charts highlight feature sub-groups that encode similar information (and their respective similarities) allowing for a principled and interpretable choice of features for classification and analysis. Using multiple electromyographic (EMG) datasets as a case study, we use this feature chart to identify functional groups among 58 state-of-the-art EMG features, and to show that they generalize across three different forearm EMG datasets obtained from able-bodied subjects during hand and finger contractions. We find that these groups describe meaningful non-redundant information, succinctly recapitulating information about different regions of feature space. We then recommend representative features from each group based on maximum class separability, robustness and minimum complexity.

Evaluation of Myoelectric Control Learning Using Multi-Session Game-Based Training

While training is critical for ensuring initial success as well as continued adoption of a myoelectric powered prosthesis, relatively little is known about the amount of training that is necessary. In previous studies, participants have completed only a small number of sessions, leaving doubt about whether the findings necessarily generalize to a longer-term clinical training program. Furthermore, a heavy emphasis has been placed on a functional prosthesis use when assessing the effectiveness of myoelectric training. Although well-motivated, this all-inclusive approach may obscure more subtle improvements made in underlying muscle control that could lead to tangible benefits. In this paper, a deeper exploration of the effects of myoelectric training was performed by following the progress of 30 participants as they completed a series of ten 30-min training sessions over multiple days. The progress was assessed using a newly developed set of metrics that was specifically designed to quantify the aspects of muscle control that are foundational to the strong myoelectric prosthesis use. It was determined that, while myoelectric training can lead to improvements in muscle control, these improvements may take longer than previously considered, even occurring after improvements in the training game itself. These results suggest the need to reconsider how and when transfer from training activities is assessed.

High-density force myography: A possible alternative for upper-limb prosthetic control

Several multiple degree-of-freedom upper-limb prostheses that have the promise of highly dexterous control have recently been developed. Inadequate controllability, however, has limited adoption of these devices. Introducing more robust control methods will likely result in higher acceptance rates. This work investigates the suitability of using high-density force myography (HD-FMG) for prosthetic control. HD-FMG uses a high-density array of pressure sensors to detect changes in the pressure patterns between the residual limb and socket caused by the contraction of the forearm muscles. In this work, HD-FMG outperforms the standard electromyography (EMG)-based system in detecting different wrist and hand gestures. With the arm in a fixed, static position, eight hand and wrist motions were classified with 0.33% error using the HD-FMG technique. Comparatively, classification errors in the range of 2.2%-11.3% have been reported in the literature for multichannel EMG-based approaches. As with EMG, position variation in HD-FMG can introduce classification error, but incorporating position variation into the training protocol reduces this effect. Channel reduction was also applied to the HD-FMG technique to decrease the dimensionality of the problem as well as the size of the sensorized area. We found that with informed, symmetric channel reduction, classification error could be decreased to 0.02%.

On the robustness of EMG features for pattern recognition based myoelectric control: a multi-dataset comparison

The selection of optimal features has long been a subject of debate for pattern recognition based myoelectric control. Studies have compared many features, but have typically used small or constrained data sets. Herein, the performance of various features is evaluated using data from six previously reported data sets. The number of channels, the contraction dynamics (dynamic vs static), and classifier type all yielded significant interactions (p<;0.05) with the feature set. When using 8 channels, the addition of the tested features produced no improvement over a standard time domain (TD) feature set alone (p>0.05). When using fewer channels, however, autoregressive, Cepstral coefficients, Willison amplitude and sample entropy features all provided significant improvement during dynamic contractions (p<;0.05). The simple Willison amplitude is highlighted, showing that it can provide significant improvement when used as a replacement for any one of the standard TD features.

A characterization of the effect of limb position on EMG features to guide the development of effective prosthetic control schemes

Electromyogram (EMG) pattern recognition has long been used for the control of upper limb prostheses. More recently, it has been shown that variability induced during functional use, such as changes in limb position and dynamic contractions, can have a substantial impact on the robustness of EMG pattern recognition. This work further investigates the reasons for pattern recognition performance degradation due to the limb position variation. The main focus is on the impact of limb position variation on features of the EMG, as measured using separability and repeatability metrics. The results show that when the limb is moved to a position different from the one in which the classifier is trained, both the separability and repeatability of the data decrease. It is shown how two previously proposed classification methods, multiple position training and dual-stage classification, resolve the position effect problem to some extent through increasing either separability or repeatability but not both. A hybrid classification method which exhibits a compromise between separability and repeatability is proposed in this work. It is shown that, when tested with the limb in 16 different positions, this method increases classification accuracy from an average of 70% (single position training) to 89% (hybrid approach). This hybrid method significantly (p<;0.05) outperforms multiple position training (an average of 86%) and dual-stage classification (an average of 85%).

A comparison of classification based confidence metrics for use in the design of myoelectric control systems

In many pattern recognition applications, confidence scores are used to extract more information than discrete class membership alone, yet they have not traditionally been leveraged in myoelectric control. In this work, the confidence scores of eight common classification schemes were examined. Their role in rejecting inadvertent motions is investigated, and the tradeoffs observed in the design of rejection capable control schemes are demonstrated. It is shown that the distribution of confidences can varying greatly between classifiers, even when classification performance is similar. As a specific example, an ensemble of support vector machines in a one against one configuration (SVM1vs1) outperforms the previously reported LDAR myoelectric pattern recognition rejection scheme in terms of accuracy-rejection curves (ARC) and false acceptance/rejection (FAR) curves.

Proceedings of the first workshop on Peripheral Machine Interfaces: going beyond traditional surface electromyography

One of the hottest topics in rehabilitation robotics is that of proper control of prosthetic devices. Despite decades of research, the state of the art is dramatically behind the expectations. To shed light on this issue, in June, 2013 the first international workshop on Present and future of non-invasive peripheral nervous system (PNS)-Machine Interfaces (MI; PMI) was convened, hosted by the International Conference on Rehabilitation Robotics. The keyword PMI has been selected to denote human-machine interfaces targeted at the limb-deficient, mainly upper-limb amputees, dealing with signals gathered from the PNS in a non-invasive way, that is, from the surface of the residuum. The workshop was intended to provide an overview of the state of the art and future perspectives of such interfaces; this paper represents is a collection of opinions expressed by each and every researcher/group involved in it.

Whitening of the electromyogram for improved classification accuracy in prosthesis control

The electromyogram (EMG) signal has been used as the command input to myoelectric prostheses. A common control scheme is based on classifying the EMG signals from multiple electrodes into one of several distinct classes of user intent/function. In this work, we investigated the use of EMG whitening as a preprocessing step to EMG pattern recognition. Whitening is known to decorrelate the EMG signal, with improved performance shown in the related applications of EMG amplitude estimation and EMG-torque processing. We reanalyzed the EMG signals recorded from 10 electrodes placed circumferentially around the forearm of 10 intact subjects and 5 amputees. The coefficient of variation of two time-domain features--mean absolute value and signal length--was significantly reduced after whitening. Pre-whitened classification models using these features, along with autoregressive power spectrum coefficients, added approximately five percentage points to their classification accuracy. Improvement was best using smaller window durations (<100 ms).

Electromyogram whitening for improved classification accuracy in upper limb prosthesis control

Time and frequency domain features of the surface electromyogram (EMG) signal acquired from multiple channels have frequently been investigated for use in controlling upper-limb prostheses. A common control method is EMG-based motion classification. We propose the use of EMG signal whitening as a preprocessing step in EMG-based motion classification. Whitening decorrelates the EMG signal and has been shown to be advantageous in other EMG applications including EMG amplitude estimation and EMG-force processing. In a study of ten intact subjects and five amputees with up to 11 motion classes and ten electrode channels, we found that the coefficient of variation of time domain features (mean absolute value, average signal length and normalized zero crossing rate) was significantly reduced due to whitening. When using these features along with autoregressive power spectrum coefficients, whitening added approximately five percentage points to classification accuracy when small window lengths were considered.

Motion Normalized Proportional Control for Improved Pattern Recognition-Based Myoelectric Control

This paper describes two novel proportional control algorithms for use with pattern recognition-based myoelectric control. The systems were designed to provide automatic configuration of motion-specific gains and to normalize the control space to the user's usable dynamic range. Class-specific normalization parameters were calculated using data collected during classifier training and require no additional user action or configuration. The new control schemes were compared to the standard method of deriving proportional control using a one degree of freedom Fitts' law test for each of the wrist flexion/extension, wrist pronation/supination and hand close/open degrees of freedom. Performance was evaluated using the Fitts' law throughput value as well as more descriptive metrics including path efficiency, overshoot, stopping distance and completion rate. The proposed normalization methods significantly outperformed the incumbent method in every performance category for able bodied subjects (p < 0.001) and nearly every category for amputee subjects. Furthermore, one proposed method significantly outperformed both other methods in throughput (p < 0.0001), yielding 21% and 40% improvement over the incumbent method for amputee and able bodied subjects, respectively. The proposed control schemes represent a computationally simple method of fundamentally improving myoelectric control users' ability to elicit robust, and controlled, proportional velocity commands.

A multi-modal approach for hand motion classification using surface EMG and accelerometers

For decades, electromyography (EMG) has been used for diagnostics, upper-limb prosthesis control, and recently even for more general human-machine interfaces. Current commercial upper limb prostheses usually have only two electrode sites due to cost and space limitations, while researchers often experiment with multiple sites. Micro-machined inertial sensors are gaining popularity in many commercial and research applications where knowledge of the postures and movements of the body is desired. In the present study, we have investigated whether accelerometers, which are relatively cheap, small, robust to noise, and easily integrated in a prosthetic socket; can reduce the need for adding more electrode sites to the prosthesis control system. This was done by adding accelerometers to a multifunction system and also to a simplified system more similar to current commercially available prosthesis controllers, and assessing the resulting changes in classification accuracy. The accelerometer does not provide information on muscle force like EMG electrodes, but the results show that it provides useful supplementary information. Specifically, if one wants to improve a two-site EMG system, one should add an accelerometer affixed to the forearm rather than a third electrode.

Improving myoelectric pattern recognition positional robustness using advanced training protocols

The control of powered upper limb prostheses using the surface electromyogram (EMG) is an important clinical option for amputees. There have been considerable recent improvements in prosthetic hands, but these currently lack a control scheme that can decode movement intent from the EMG to exploit their mechanical dexterity. Pattern recognition based control has the potential to decode many classes of movement intent, but is confounded when using the prosthesis in varying positions during activities of daily living. This work describes the degradation that can occur when using pattern recognition in varying positions, during both static positioning tasks and dynamic activities of daily living. It is shown that training with dynamic activities can greatly improve positional robustness for both static and dynamic tasks, without requiring a complex and lengthy training session.