Prediction of gait recovery using machine learning algorithms in patients with spinal cord injury.

With advances in artificial intelligence, machine learning (ML) has been widely applied to predict functional outcomes in clinical medicine. However, there has been no attempt to predict walking ability after spinal cord injury (SCI) based on ML. In this situation, the main purpose of this study was to predict gait recovery after SCI at discharge from an acute rehabilitation facility using various ML algorithms. In addition, we explored important variables that were related to the prognosis. Finally, we attempted to suggest an ML-based decision support system (DSS) for predicting gait recovery after SCI. Data were collected retrospectively from patients with SCI admitted to an acute rehabilitation facility between June 2008 to December 2021. Linear regression analysis and ML algorithms (random forest [RF], decision tree [DT], and support vector machine) were used to predict the functional ambulation category at the time of discharge (FAC_DC) in patients with traumatic or non-traumatic SCI (n = 353). The independent variables were age, sex, duration of acute care and rehabilitation, comorbidities, neurological information entered into the International Standards for Neurological Classification of SCI worksheet, and somatosensory-evoked potentials at the time of admission to the acute rehabilitation facility. In addition, the importance of variables and DT-based DSS for FAC_DC was analyzed. As a result, RF and DT accurately predicted the FAC_DC measured by the root mean squared error. The root mean squared error of RF and the DT were 1.09 and 1.24 for all participants, 1.20 and 1.06 for those with trauma, and 1.12 and 1.03 for those with non-trauma, respectively. In the analysis of important variables, the initial FAC was found to be the most influential factor in all groups. In addition, we could provide a simple DSS based on strong predictors such as the initial FAC, American Spinal Injury Association Impairment Scale grades, and neurological level of injury. In conclusion, we provide that ML can accurately predict gait recovery after SCI for the first time. By focusing on important variables and DSS, we can guide early prognosis and establish personalized rehabilitation strategies in acute rehabilitation hospitals.

Exploring role of microbatteries in enhancing sustainability and functionality of implantable biosensors and bioelectronics.

Microbatteries are emerging as a sustainable, miniaturized power source, crucial for implantable biomedical devices. Their significance lies in offering high energy density, longevity, and rechargeability, facilitating uninterrupted health monitoring and treatment within the body. The review delves into the development of microbatteries, emphasizing their miniaturization and biocompatibility, crucial for long-term, safe in-vivo use. It examines cutting-edge manufacturing techniques like physical and chemical vapor deposition, and atomic layer deposition, essential for the precision manufacture of the microbatteries. The paper contrasts primary and secondary batteries, highlighting the advantages of zinc-ion and magnesium-ion batteries for enhanced stability and reduced reactivity. It also explores biodegradable batteries, potentially obviating the need for surgical extraction post-use. The integration of microbatteries into diagnostic and therapeutic devices is also discussed, illustrating how they enhance the efficacy and sustainability of implantable biosensors and bioelectronics.

Cobalt MOF-Based Porous Carbonaceous Spheres for Multimodal Soft Actuator Exhibiting Intricate Biomimetic Motions.

The advancement of active electrode materials is essential to meet the demand for multifaceted soft robotic interactions. In this study, a new type of porous carbonaceous sphere (PCS) for a multimodal soft actuator capable of both magnetoactive and electro-ionic responses is reported. The PCS, derived from the simultaneous oxidative and reductive breakdown of specially designed cobalt-based metal-organic frameworks (Co-MOFs) with varying metal-to-ligand ratios, exhibits a high specific surface area of 529 m2 g-1 and a saturated magnetization of 142.7 Am2 kg-1. The size of the PCS can be controlled through the Ostwald ripening mechanism, while the porous structure can be regulated by adjusting the metal-to-ligand mol ratio. Its exceptional compatibility with poly(3,4-ethylene-dioxythiophene)-poly(styrenesulfonate) enables the creation of uniform electrode, crucial for producing soft actuators that work in both magnetic and electrical fields. Operated at an ultralow voltage of 1 V, the PCS-based actuator generates a blocking force of 47.5 mN and exhibits significant bending deflection even at an oscillation frequency of 10 Hz. Employing this simultaneous multimodal actuation ensures the dynamic and complex motions of a balancing bird robot and a dynamic eagle robot. This advancement marks a significant step toward the realization of more dynamic and versatile soft robotic systems.

Influence of MXene Composition on Triboelectricity of MXene-Alginate Nanocomposites.

MXenes are highly versatile and conductive 2D materials that can significantly enhance the triboelectric properties of polymer nanocomposites. Despite the growing interest in the tunable chemistry of MXenes for energy applications, the effect of their chemical composition on triboelectric power generation has yet to be thoroughly studied. Here, we investigate the impact of the chemical composition of MXenes, specifically the Ti3CNTx carbonitride vs the most studied carbide, Ti3C2Tx, on their interactions with sodium alginate biopolymer and, ultimately, the performance of a triboelectric nanogenerator (TENG) device. Our results show that adding 2 wt % of Ti3CNTx to alginate produces a synergistic effect that generates a higher triboelectric output than the Ti3C2Tx system. Spectroscopic analyses suggest that a higher oxygen and fluorine content on the surface of Ti3CNTx enhances hydrogen bonding with the alginate matrix, thereby increasing the surface charge density of the alginate oxygen atoms. This was further supported by Kelvin probe force microscopy, which revealed a more negative surface potential on Ti3CNTx-alginate, facilitating high charge transfer between the TENG electrodes. The optimized Ti3CNTx-alginate nanogenerator delivered an output of 670 V, 15 µA, and 0.28 W/m2. Additionally, we demonstrate that plasma oxidation of the MXene surface further enhances triboelectric performance. Due to the diverse surface terminations of MXene, we show that Ti3CNTx-alginate can function as either tribopositive or tribonegative material, depending on the counter-contacting material. Our findings provide a deeper understanding of how MXene composition affects their interaction with biopolymers and resulting tunable triboelectrification behavior. This opens up new avenues for developing flexible and efficient MXene-based TENG devices.

Deep Learning-Based Prediction Model for Gait Recovery after a Spinal Cord Injury.

Predicting gait recovery after a spinal cord injury (SCI) during an acute rehabilitation phase is important for planning rehabilitation strategies. However, few studies have been conducted on this topic to date. In this study, we developed a deep learning-based prediction model for gait recovery after SCI upon discharge from an acute rehabilitation facility. Data were collected from 405 patients with acute SCI admitted to the acute rehabilitation facility of Korea University Anam Hospital between June 2008 and December 2022. The dependent variable was Functional Ambulation Category at the time of discharge (FAC-DC). Seventy-one independent variables were selected from the existing literature: basic information, International Standards for Neurological Classification of SCI scores, neurogenic bladders, initial FAC, and somatosensory-evoked potentials of the lower extremity. Recurrent neural network (RNN), linear regression (LR), Ridge, and Lasso methods were compared for FAC-DC prediction in terms of the root-mean-squared error (RMSE). RNN variable importance, which is the RMSE gap between a complete RNN model and an RNN model excluding a certain variable, was used to evaluate the contribution of this variable. Based on the results of this study, the performance of the RNN was far better than that of LR, Ridge, and Lasso. The respective RMSEs were 0.3738, 2.2831, 1.3161, and 1.0246 for all the participants; 0.3727, 1.7176, 1.3914, and 1.3524 for those with trauma; and 0.3728, 1.7516, 1.1012, and 0.8889 for those without trauma. In terms of RNN variable importance, lower-extremity motor strength (right and left ankle dorsiflexors, right knee extensors, and left long toe extensors) and the neurological level of injury were ranked among the top five across the boards. Therefore, initial FAC was the seventh, third, and ninth most important predictor for all participants, those with trauma, and those without trauma, respectively. In conclusion, this study developed a deep learning-based prediction model with excellent performance for gait recovery after SCI at the time of discharge from an acute rehabilitation facility. This study also demonstrated the strength of deep learning as an explainable artificial intelligence method for identifying the most important predictors.

Janus CoMOF-SEBS Membrane for Bifunctional Dielectric Layer in Triboelectric Nanogenerators.

Considerable research has been conducted on the application of functional nano-fillers to enhance the power generation capabilities of triboelectric nanogenerators (TENGs). However, these additives often exhibit a decrease in output power at higher concentration. Here, a Janus cobalt metal-organic framework-SEBS (JCMS) membrane is reported as a dual-purpose dielectric layer capable of efficiently capturing and blocking charges for high-performance TENGs. The JCMS is produced asymmetrically through gravitational sedimentation, employing spherical CoMOFs within a diluted SEBS solution. Beyond its dual dielectric characteristics, the JCMS showcases exceptional mechanical durability, displaying notable stretchability of up to 475% and remarkable resilience when subjected to diverse mechanical pressures. Consequently, the JCMS-TENG produces a maximum peak-to-peak voltage of 936 V, a current of 42.8 µA, and a power density of 10.89 W m- 2 when exposed to an external force of 10 N at a 5 Hz frequency. This investigation highlights the potential of JCMS-TENGs with unique structures, known for their exceptional energy harvesting capabilities, mechanical strength, and flexibility. Additionally, the promising prospects of easily produced asymmetric structures is emphasized with bifunctionalities for developing efficient and flexible MOFs-based TENGs. These advancements are well-suited for self-powered wearables, rehabilitation devices, and energy harvesters.

Functionally antagonistic polyelectrolyte for electro-ionic soft actuator.

Electro-active ionic soft actuators have been intensively investigated as an artificial muscle for soft robotics due to their large bending deformations at low voltages, small electric power consumption, superior energy density, high safety and biomimetic self-sensing actuation. However, their slow responses, poor durability and low bandwidth, mainly resulting from improper distribution of ionic conducting phase in polyelectrolyte membranes, hinder practical applications to real fields. We report a procedure to synthesize efficient polyelectrolyte membranes that have continuous conducting network suitable for electro-ionic artificial muscles. This functionally antagonistic solvent procedure makes amphiphilic Nafion molecules to assemble into micelles with ionic surfaces enclosing non-conducting cores. Especially, the ionic surfaces of these micelles combine together during casting process and form a continuous ionic conducting phase needed for high ionic conductivity, which boosts the performance of electro-ionic soft actuators by 10-time faster response and 36-time higher bending displacement. Furthermore, the developed muscle shows exceptional durability over 40 days under continuous actuation and broad bandwidth below 10 Hz, and is successfully applied to demonstrate an inchworm-mimetic soft robot and a kinetic tensegrity system.

Automated non-contact measurement of the spine curvature at the sagittal plane using a deep neural network.

BACKGROUND: Non-radiographical techniques have been suggested to measure the spine curvature at the sagittal plane. However, a neural network has not been used to measure the curvature. METHODS: A single video camera captured images of a standing posture at the sagittal plane from twenty healthy males. Six marker positions along the spine's contour in each image were identified for measuring inclination, thoracic kyphosis, and lumbar lordosis angles. We estimated three inflection points around the neck, hip, and between the neck and hip, followed by identifying two adjacent marker positions per inflection point to compute its tangent. The angular deviation of each tangent line from the horizontal was computed to measure inclination angles. Thoracic kyphosis and lumbar lordosis angles were computed by the angular difference between the two adjacent tangents. A deep neural network was trained with 500,000 iterations using the labeled images from 18 participants (388 and 44 images for training and test set) and then evaluated using the unseen images (2 participants, 48 images; evaluation set). FINDINGS: The mean total training and test errors were <2 pixels (∼ 0.6 cm). The total error in the evaluation set was qualitatively comparable (∼ 3 pixels = âˆ¼ 0.9 cm), suggesting the model performance was maintained in the unseen data. The angle values between labeled and network-predicted marker positions were similar in the evaluation set. INTERPRETATION: The network training with a relatively small number of images was successful based on the small error values observed in the evaluation set. The model may be an affordable, automated, and non-contact measurement tool for the human spine curvature.

Polysulfonated covalent organic framework as active electrode host for mobile cation guests in electrochemical soft actuator.

Tailoring transfer dynamics of mobile cations across solid-state electrolyte-electrode interfaces is crucial for high-performance electrochemical soft actuators. In general, actuation performance is directly proportional to the affinity of cations and anions in the electrolyte for the opposite electrode surfaces under an applied field. Herein, to maximize electrochemical actuation, we report an electronically conjugated polysulfonated covalent organic framework (pS-COF) used as a common electrolyte-electrode host for 1-ethyl-3-methylimidazolium cation embedded into a Nafion membrane. The pS-COF-based electrochemical actuator exhibits remarkable bending deflection at near-zero voltage (~0.01 V) and previously unattainable blocking force, which is 34 times higher than its own weight. The ultrafast step response shows a very short rising time of 1.59 seconds without back-relaxation, and substantial ultralow-voltage actuation at higher frequencies up to 5.0 hertz demonstrates good application prospects of common electrolyte-electrode hosts. A soft fluidic switch is constructed using the proposed soft actuator as a potential engineering application.

Easy-To-Wear Auxetic SMA Knot-Architecture for Spatiotemporal and Multimodal Haptic Feedbacks.

Wearable haptic interfaces prioritize user comfort, but also value the ability to provide diverse feedback patterns for immersive interactions with the virtual or augmented reality. Here, to provide both comfort and diverse tactile feedback, an easy-to-wear and multimodal wearable haptic auxetic fabric (WHAF) is prepared by knotting shape-memory alloy wires into an auxetic-structured fabric. This unique meta-design allows the WHAF to completely expand and contract in 3D, providing superior size-fitting and shape-fitting capabilities. Additionally, a microscale thin layer of Parylene is coated on the surface to create electrically separated zones within the WHAF, featuring zone-specified actuation for conveying diverse spatiotemporal information to users with using the WHAF alone. Depending on the body part it is worn on, the WHAF conveys either cutaneous or kinesthetic feedback, thus, working as a multimodal wearable haptic interface. As a result, when worn on the forearm, the WHAF intuitively provides spatiotemporal information to users during hands-free navigation and teleoperation in virtual reality, and when worn on the elbow, the WHAF guides users to reach the desired elbow flexion, like a personal exercise advisor.

Magnetically and Electrically Responsive Soft Actuator Derived from Ferromagnetic Bimetallic Organic Framework.

The advancement in smart devices and soft robotics necessitates the use of multiresponsive soft actuators with high actuation stroke and stable reversibility for their use in real-world applications. Here, this work reports a magnetically and electrically dual responsive soft actuator based on neodymium and iron bimetallic organic frameworks (NdFeMOFs@700). The ferromagnetic NdFeMOFs@700 exhibits a porous carbon structure with excellent magnetization saturation (166.96 emu g-1 ) which allows its application to a dual functional material in both magnetoactive and electro-ionic actuations. The electro-ionic soft actuator, which is fabricated using NdFeMOFs@700 and PEDOT-PSS, demonstrates 4.5 times higher ionic charge storage capacity (68.21 mF cm-2 ) and has excellent cycle stability compared with the PEDOT-PSS based actuator. Under a low sinusoidal input voltage of 1 V, the dual-responsive actuator displays bending displacement of 15.46 mm and also generates deflection of 10 mm at 50 mT. Present results show that the ferromagnetic bimetallic organic frameworks can open a new way to make dual responsive soft actuators due to the hierarchically porous structures with its high redox activity, superior magnetic properties, and larger electrochemical capacitance. With the NdFeMOFs@700 based soft actuators, walking movement of a starfish robot is demonstrated by applying both the magnetic and electric fields.

Clinical efficacy of overground powered exoskeleton for gait training in patients with subacute stroke: A randomized controlled pilot trial.

BACKGROUND: To investigate the efficacy and usefulness of 12 sessions of overground robot-assisted gait training (RAGT) in subacute stroke patients. METHODS: In this pilot study, 17 subacute stroke survivors were randomly assigned to the intervention (n = 9) and control (n = 8) groups. In addition to the conventional stroke neurorehabilitation program, the intervention group received 30 minutes of overground exoskeletal RAGT, while the control group received 30 minutes of conventional gait training by a physiotherapist. All interventions were performed in 12 sessions (3 times/week for 4 weeks). The primary aim was to assess ambulation ability using the functional ambulation category (FAC). The 10-m walk test, Berg Balance Scale, timed-up-and-go Timed-up-and-go, Fugl-Meyer assessment of lower extremity, pulmonary function test, the Korean version of the modified Barthel index, and Euro quality of life-5 dimensions (EQ-5D) were assessed. All outcomes were evaluated both before and after the intervention. RESULTS: The Berg Balance Scale, Korean version of the modified Barthel index, and EQ-5D scores (P < .05) improved significantly in both groups. Only those in the RAGT group improved significantly in the FAC, timed-up-and-go, and 10-m walk test (P < .05). In the FAC and EQ-5D, the intervention group showed greater improvement than the control group (P < .05). CONCLUSION: We found that 4 weeks of overground RAGT combined with conventional training may improve walking independence and quality of life in patients with subacute stroke.

Effects of electrical muscle stimulation on core muscle activation and physical performance in non-athletic adults: A randomized controlled trial.

BACKGROUND: Electrical muscle stimulation (EMS) activates muscles through electrical currents, resulting in involuntary muscle contractions. This study aimed to evaluate the immediate clinical effects of superimposing EMS on strength training compared with conventional exercise in healthy non-athletic adults. METHODS: This study was a randomised, controlled, parallel-group trial conducted at a single centre. Forty-one healthy young volunteers were recruited and randomised into two groups: strengthening with superimposed EMS (S+E) and strengthening (S) groups. All participants underwent the 30 minutes of strength training program, three times a week for 8 weeks, consisting of core muscle exercises. Additionally, the S+E group received EMS during training, which stimulated the bilateral abdominal, gluteus, and hip adductor muscles. As the primary outcome measure, we evaluated the changes in muscle thickness, including the abdominal, gluteal, and hip adductor muscles, using ultrasound. Muscle thickness was measured in both resting and contracted states. For secondary outcomes, physical performance (Functional Movement System score, McGill's core stability test, and hip muscle power) and body composition analysis were evaluated. All assessments were performed at the beginning and end of the intervention. RESULTS: 39 participants (S+E group = 20, S group = 19) completed the study. The clinical characteristics and baseline functional status of each group did not differ significantly between the groups. After completion of the training, the S+E group showed more efficient contraction in most of the evaluated muscles. The resting muscle thickness did not differ significantly between the groups; however, the contracted muscle thickness in the S+E group was higher than that in the S group (p < 0.05). Physical performance and body composition were not significantly different between the two groups. No intervention-related complications were reported during the study. CONCLUSION: EMS seems to be a safe and reasonable modality for improving physical fitness in healthy individuals.

Vowel speech recognition from rat electroencephalography using long short-term memory neural network.

Over the years, considerable research has been conducted to investigate the mechanisms of speech perception and recognition. Electroencephalography (EEG) is a powerful tool for identifying brain activity; therefore, it has been widely used to determine the neural basis of speech recognition. In particular, for the classification of speech recognition, deep learning-based approaches are in the spotlight because they can automatically learn and extract representative features through end-to-end learning. This study aimed to identify particular components that are potentially related to phoneme representation in the rat brain and to discriminate brain activity for each vowel stimulus on a single-trial basis using a bidirectional long short-term memory (BiLSTM) network and classical machine learning methods. Nineteen male Sprague-Dawley rats subjected to microelectrode implantation surgery to record EEG signals from the bilateral anterior auditory fields were used. Five different vowel speech stimuli were chosen, /a/, /e/, /i/, /o/, and /u/, which have highly different formant frequencies. EEG recorded under randomly given vowel stimuli was minimally preprocessed and normalized by a z-score transformation to be used as input for the classification of speech recognition. The BiLSTM network showed the best performance among the classifiers by achieving an overall accuracy, f1-score, and Cohen's κ values of 75.18%, 0.75, and 0.68, respectively, using a 10-fold cross-validation approach. These results indicate that LSTM layers can effectively model sequential data, such as EEG; hence, informative features can be derived through BiLSTM trained with end-to-end learning without any additional hand-crafted feature extraction methods.

A Dual-Responsive Magnetoactive and Electro-Ionic Soft Actuator Derived from a Nickel-Based Metal-Organic Framework.

There is growing demand for multiresponsive soft actuators for the realization of natural, safe, and complex motions in robotic interactions. In particular, soft actuators simultaneously stimulated by electrical and magnetic fields are always under development owing to their simple controllability and reliability during operation. Herein, magnetically and electrically driven dual-responsive soft actuators (MESAs) derived from novel nickel-based metal-organic frameworks (Ni-MOFs-700C), are reported. Nanoscale Ni-MOFs-700C has excellent electrochemical and magnetic properties that allow it to be used as a multifunctional material under both magnetoactive and electro-ionic actuations. The dual-responsive MESA exhibits a bending displacement of 30 mm and an ultrafast rising time of 1.5 s under a very low input voltage of 1 V and also exerts a bending deflection of 12.5 mm at 50 mT under a high excitation frequency of 5 Hz. By utilizing a dual-responsive MESA, the hovering motion of a hummingbird robot is demonstrated under magnetic and electrical stimuli.

Fabrication of Three-Dimensional Multilayer Structures of Single-Walled Carbon Nanotubes Based on the Plasmonic Carbonization.

We developed a complex three-dimensional (3D) multilayer deposition method for the fabrication of single-walled carbon nanotubes (SWCNTs) using vacuum filtration and plasmonic carbonization without lithography and etching processes. Using this fabrication method, SWCNTs can be stacked to form complex 3D structures that have a large surface area relative to the unit volume compared to the single-plane structure of conventional SWCNTs. We characterized 3D multilayer SWCNT patterns using a surface optical profiler, Raman spectroscopy, sheet resistance, scanning electron microscopy, and contact angle measurements. Additionally, these carbon nanotube (CNT) patterns showed excellent mechanical stability even after elastic bending tests more than 1000 times at a radius of 2 mm.

Quantification of stroke lesion volume using epidural EEG in a cerebral ischaemic rat model.

Precise monitoring of the brain after a stroke is essential for clinical decision making. Due to the non-invasive nature and high temporal resolution of electroencephalography (EEG), it is widely used to evaluate real-time cortical activity. In this study, we investigated the stroke-related EEG biomarkers and developed a predictive model for quantifying the structural brain damage in a focal cerebral ischaemic rat model. We enrolled 31 male Sprague-Dawley rats and randomly assigned them to mild stroke, moderate stroke, severe stroke, and control groups. We induced photothrombotic stroke targeting the right auditory cortex. We then acquired EEG signal responses to sound stimuli (frequency linearly increasing from 8 to 12 kHz with 750 ms duration). Power spectral analysis revealed a significant correlation of the relative powers of alpha, theta, delta, delta/alpha ratio, and (delta + theta)/(alpha + beta) ratio with the stroke lesion volume. The auditory evoked potential analysis revealed a significant association of amplitude and latency with stroke lesion volume. Finally, we developed a multiple regression model combining EEG predictors for quantifying the ischaemic lesion (R2 = 0.938, p value < 0.001). These findings demonstrate the potential application of EEG as a valid modality for monitoring the brain after a stroke.

Transient lower extremity lymphedema following COVID-19 vaccination: A case report.

RATIONALE: Complications from COVID-19 vaccines have yet to be sufficiently analyzed because they are rapidly approved without long-term data. In particular, there are no case reports of lymphedema in a healthy patient following vaccination. Herein, we report a patient who underwent transient lymphedema after vaccination with BNT16b2. PATIENT CONCERNS: A 79-year-old woman with pitting edema in both lower legs after administration of a second dose of Pfizer vaccine was referred to our clinic. In the absence of clinical evidence of swelling during the laboratory evaluation, we suspected deep vein thrombosis. However, ultrasonographic findings revealed no evidence of venous thrombosis or varicose veins. DIAGNOSIS: On the basis of lymphoscintigraphy, the patient was diagnosed with transient lymphedema with decreased lymphatic transport in both lower extremities. INTERVENTION: The patient received intensive physiotherapy, including complex decongestive physiotherapy and pneumatic pump compression, to improve the lymphatic circulation. Furthermore, the patient was trained to apply a multilayer compressive bandage to the lower extremities. OUTCOMES: At 2 months follow-up after rehabilitative treatment, the patient's symptoms improved without recurring lymphedema. LESSONS: In the absence of clinical evidence of swelling during laboratory evaluation or ultrasonographic investigations suggesting deep vein thrombosis, we should consider the possibility of lymphatic disorders.

Myoelectric Signal Classification of Targeted Muscles Using Dictionary Learning.

Surface electromyography (sEMG) signals comprise electrophysiological information related to muscle activity. As this signal is easy to record, it is utilized to control several myoelectric prostheses devices. Several studies have been conducted to process sEMG signals more efficiently. However, research on optimal algorithms and electrode placements for the processing of sEMG signals is still inconclusive. In addition, very few studies have focused on minimizing the number of electrodes. In this study, we investigated the most effective method for myoelectric signal classification with a small number of electrodes. A total of 23 subjects participated in the study, and the sEMG data of 14 different hand movements of the subjects were acquired from targeted muscles and untargeted muscles. Furthermore, the study compared the classification accuracy of the sEMG data using discriminative feature-oriented dictionary learning (DFDL) and other conventional classifiers. DFDL demonstrated the highest classification accuracy among the classifiers, and its higher quality performance became more apparent as the number of channels decreased. The targeted method was superior to the untargeted method, particularly when classifying sEMG signals with DFDL. Therefore, it was concluded that the combination of the targeted method and the DFDL algorithm could classify myoelectric signals more effectively with a minimal number of channels.

Gesture Classification from Compressed EMG Based on Compressive Covariance Sensing.

Electromyogram (EMG) based human computer interface (HCI) is an attractive technique to monitor a patient, control an artificial arm, or play a game. Since EMG processing requires high sampling and transmission rates, a compression technique is important to implement an ultra-low power wireless EMG system. Previous study has a limitation due to the complexity of algorithm and the non-sparsity nature of EMG. In this study, we proposed a new EMG compression scheme based on a compressive covariance sensing (CCS). The covariance recovered from compressed EMG was used to classify user's gestures. The proposed method was verified with NinaPro open source data, which contains 49 gestures with 6 times repetition. As a result, the proposed CCS based EMG compression technique showed good covariance recovery performance and high classification rate as well as superior compression rate.