Dataset Information

Supervised Myoelectrical Hand Gesture Recognition in Post-Acute Stroke Patients with Upper Limb Paresis on Affected and Non-Affected Sides.

ABSTRACT: In clinical practice, acute post-stroke paresis of the extremities fundamentally complicates timely rehabilitation of motor functions; however, recently, residual and distorted musculoskeletal signals have been used to initiate feedback-driven solutions for establishing motor rehabilitation. Here, we investigate the possibilities of basic hand gesture recognition in acute stroke patients with hand paresis using a novel, acute stroke, four-component multidomain feature set (ASF-4) with feature vector weight additions (ASF-14NP, ASF-24P) and supervised learning algorithms trained only by surface electromyography (sEMG). A total of 19 (65.9 ± 12.4 years old; 12 men, seven women) acute stroke survivors (12.4 ± 6.3 days since onset) with hand paresis (Brunnstrom stage 4 ± 1/4 ± 1, SIAS 3 ± 1/3 ± 2, FMA-UE 40 ± 20) performed 10 repetitive hand movements reflecting basic activities of daily living (ADLs): rest, fist, pinch, wrist flexion, wrist extension, finger spread, and thumb up. Signals were recorded using an eight-channel, portable sEMG device with electrode placement on the forearms and thenar areas of both limbs (four sensors on each extremity). Using data preprocessing, semi-automatic segmentation, and a set of extracted feature vectors, support vector machine (SVM), linear discriminant analysis (LDA), and k-nearest neighbors (k-NN) classifiers for statistical comparison and validity (paired t-tests, p-value < 0.05), we were able to discriminate myoelectrical patterns for each gesture on both paretic and non-paretic sides. Despite any post-stroke conditions, the evaluated total accuracy rate by the 10-fold cross-validation using SVM among four-, five-, six-, and seven-gesture models were 96.62%, 94.20%, 94.45%, and 95.57% for non-paretic and 90.37%, 88.48%, 88.60%, and 89.75% for paretic limbs, respectively. LDA had competitive results using PCA whereas k-NN was a less efficient classifier in gesture prediction. Thus, we demonstrate partial efficacy of the combination of sEMG and supervised learning for upper-limb rehabilitation procedures for early acute stroke motor recovery and various treatment applications.

SUBMITTER: Anastasiev A

PROVIDER: S-EPMC9692557 | biostudies-literature | 2022 Nov

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Supervised Myoelectrical Hand Gesture Recognition in Post-Acute Stroke Patients with Upper Limb Paresis on Affected and Non-Affected Sides.

Anastasiev Alexey A Kadone Hideki H Marushima Aiki A Watanabe Hiroki H Zaboronok Alexander A Watanabe Shinya S Matsumura Akira A Suzuki Kenji K Matsumaru Yuji Y Ishikawa Eiichi E

Sensors (Basel, Switzerland) 20221111 22

In clinical practice, acute post-stroke paresis of the extremities fundamentally complicates timely rehabilitation of motor functions; however, recently, residual and distorted musculoskeletal signals have been used to initiate feedback-driven solutions for establishing motor rehabilitation. Here, we investigate the possibilities of basic hand gesture recognition in acute stroke patients with hand paresis using a novel, acute stroke, four-component multidomain feature set (ASF-4) with feature ve ...[more]

PMID: 36433330

Similar Datasets

Project description:Repeated injection cycles with abobotulinumtoxinA, a botulinum toxin type A, are recommended in current clinical guidelines as a treatment option for adults with upper limb spastic paresis. However, the magnitude of the maximal therapeutic effect of repeated abobotulinumtoxinA treatment across different efficacy parameters and the number of injection cycles required to reach maximal effect remain to be elucidated. Here, we present a post hoc exploratory analysis of a randomized, double-blind, placebo-controlled trial (12-24 weeks; NCT01313299) and open-label extension study (up to 12 months; NCT0131331), in patients aged 18-80 years with hemiparesis for ≥6 months after stroke/traumatic brain injury. Two inferential methods were used to assess the changes in efficacy parameters after repeat abobotulinumtoxinA treatment cycles: Mixed Model Repeated Measures analysis and Non-Linear Random Coefficients analysis. Using the latter model, the expected maximal effect size (not placebo-controlled) and the number of treatment cycles to reach 90% of this maximal effect were estimated. Treatment responses in terms of passive and perceived parameters (i.e. modified Ashworth scale in primary target muscle group, disability assessment scale for principal target for treatment or limb position, and angle of catch at fast speed) were estimated to reach near-maximal effect in two to three cycles. Near-maximal treatment effect for active parameters (i.e. active range of motion against the resistance of extrinsic finger flexors and active function, assessed by the Modified Frenchay Scale) was estimated to be reached one to two cycles later. In contrast to most parameters, active function showed greater improvements at Week 12 (estimated maximal change from baseline-modified Frenchay Scale overall score: +0.8 (95% confidence interval, 0.6; 1.0) than at Week 4 (+0.6 [95% confidence interval, 0.4; 0.8]). Overall, the analyses suggest that repeated treatment cycles with abobotulinumtoxinA in patients chronically affected with upper limb spastic paresis allow them to relearn how to use the affected arm with now looser antagonists. Future studies should assess active parameters as primary outcome measures over repeated treatment cycles, and assess efficacy at the 12-week time-point of each cycle, as the benefits of abobotulinumtoxinA may be underestimated in the studies of insufficient duration. Abbreviated summary In this post hoc analysis of repeated abobotulinumtoxinA injection cycles in upper limb spastic paresis, Gracies et al. used statistical modelling to elucidate the maximal therapeutic effect of abobotulinumtoxinA. Notably, the number of injections required to reach this maximal effect was higher for active (e.g. active function) compared with passive (e.g. tone) parameters.

Project description:Hand gesture recognition tasks based on surface electromyography (sEMG) are vital in human-computer interaction, speech detection, robot control, and rehabilitation applications. However, existing models, whether traditional machine learnings (ML) or other state-of-the-arts, are limited in the number of movements. Targeting a large number of gesture classes, more data features such as temporal information should be persisted as much as possible. In the field of sEMG-based recognitions, the recurrent convolutional neural network (RCNN) is an advanced method due to the sequential characteristic of sEMG signals. However, the invariance of the pooling layer damages important temporal information. In the all convolutional neural network (ACNN), because of the feature-mixing convolution operation, a same output can be received from completely different inputs. This paper proposes a concatenate feature fusion (CFF) strategy and a novel concatenate feature fusion recurrent convolutional neural network (CFF-RCNN). In CFF-RCNN, a max-pooling layer and a 2-stride convolutional layer are concatenated together to replace the conventional simple dimensionality reduction layer. The featurewise pooling operation serves as a signal amplitude detector without using any parameter. The feature-mixing convolution operation calculates the contextual information. Complete evaluations are made on both the accuracy and convergence speed of the CFF-RCNN. Experiments are conducted using three sEMG benchmark databases named DB1, DB2 and DB4 from the NinaPro database. With more than 50 gestures, the classification accuracies of the CFF-RCNN are 88.87% on DB1, 99.51% on DB2, and 99.29% on DB4. These accuracies are the highest compared with reported accuracies of machine learnings and other state-of-the-art methods. To achieve accuracies of 86%, 99% and 98% for the RCNN, the training time are 2353.686 s, 816.173 s and 731.771 s, respectively. However, for the CFF-RCNN to reach the same accuracies, it needs only 1727.415 s, 542.245 s and 576.734 s, corresponding to a reduction of 26.61%, 33.56% and 21.19% in training time. We concluded that the CFF-RCNN is an improved method when classifying a large number of hand gestures. The CFF strategy significantly improved model performance with higher accuracy and faster convergence as compared to traditional RCNN.

Project description:BackgroundHand knob infarction is a well-known stroke entity. Based on very limited data, embolic stroke mechanism has been considered the most frequent cause; however, prognosis is considered good. We wanted to shed more light on this phenomenon by assessing a cohort of patients referred to a general hospital stroke unit.MethodsEvery subject admitted to our stroke unit with an acute isolated hand paresis in the period from 2007 to 2012 was identified prospectively. Patients who had suffered from a stroke in the hand motor cortex or an adjacent area explaining the acute loss of hand function were included in the study. The Trial of Org 10172 in Acute Stroke Treatment criteria were used to classify subtypes of stroke according to etiology. The patients were followed up during autumn 2012.ResultsSeventeen subjects were admitted, but in 2 of them symptoms were transitory and magnetic resonance imaging was negative. Two patients were excluded due to persisting sensory deficits. The remaining 13 (11 males and 2 females) patients with an average age of 62.9 (± 13.4) years were included, representing 1.5% of all ischemic strokes diagnosed at the stroke unit in the given period. All patients were right-handed, and the dominant hand was affected only in 4 (31%). The average Medical Research Council's scale score was 3.1 (± 1.4) on admission, and classified as bad. On follow-up, which occurred on average 29.8 (± 19.8) months after the stroke, the score was 4.6 (± 0.4) and was classified as fair to good. No patient experienced a new stroke. The outcome was good to excellent in 10 patients (77%). Two patients died (15%), 1 of probable cardiac arrest and 1 of unknown cause. One patient did not participate in the follow-up. The majority of patients had evidence of both small artery (77%) and large artery (85%) disease. On average, there were 1.6 (± 0.4) new ischemic lesions per patient. Six patients had a solitary lesion (46%). In 5 of them, small artery occlusion was considered the probable stroke mechanism. In 4 cases, the stroke was of undetermined etiology. Three patients had atrial fibrillation, and in 2 of them cardioembolism was the probable stroke mechanism. Two patients with definite large artery atherosclerosis underwent carotid endarterectomy, and 1 of them had comorbid atrial fibrillation.ConclusionStrokes causing isolated hand paresis seem to have a heterogeneous etiology. Prognosis regarding hand function is good, but long-term outcome depends on stroke etiology and secondary prophylaxis.

Dataset Information

Supervised Myoelectrical Hand Gesture Recognition in Post-Acute Stroke Patients with Upper Limb Paresis on Affected and Non-Affected Sides.

Publications

Supervised Myoelectrical Hand Gesture Recognition in Post-Acute Stroke Patients with Upper Limb Paresis on Affected and Non-Affected Sides.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets