Kinect-based assessment of proximal arm non-use after a stroke

Bakhti, Karima; Laffont, I.; Muthalib, Makii; Froger, Jérôme; Mottet, Denis

doi:10.1186/s12984-018-0451-2

Cited by 48 publications

(47 citation statements)

References 64 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Stroke survivors often compensate for the loss of motor function by adapting their movement patterns to incorporate additional degrees of freedom at other joints and body segments. During the seated reaching motion with their paretic upper limb, many patients spontaneously employed replace the use of their arm by recruiting excessive trunk or scapular movements, even though they are able to use their arm when forced to do so [3].…”

Section: Introductionmentioning

confidence: 99%

Detecting compensatory movements of stroke survivors using pressure distribution data and machine learning algorithms

Cai

Zhang

et al. 2019

J NeuroEngineering Rehabil

View full text Add to dashboard Cite

Background: Compensatory movements are commonly employed by stroke survivors during seated reaching and may have negative effects on their long-term recovery. Detecting compensation is useful for coaching the patient to reduce compensatory trunk movements and improving the motor function of the paretic arm. Sensor-based and camera-based systems have been developed to detect compensatory movements, but they still have some limitations, such as causing object obstructions, requiring complex setups and raising privacy concerns. To overcome these drawbacks, this paper proposes a compensatory movement detection system based on pressure distribution data and is unobtrusive, simple and practical. Machine learning algorithms were applied to classify compensatory movements automatically. Therefore, the purpose of this study was to develop and test a pressure distribution-based system for the automatic detection of compensation movements of stroke survivors using machine learning algorithms. Methods: Eight stroke survivors performed three types of reaching tasks (back-and-forth, side-to-side, and up-anddown reaching tasks) with both the healthy side and the affected side. The pressure distribution data were recorded, and five features were extracted for classification. The k-nearest neighbor (k-NN) and support vector machine (SVM) algorithms were applied to detect and categorize the compensatory movements. The surface electromyography (sEMG) signals of nine trunk muscles were acquired to provide a detailed description and explanation of compensatory movements. Results: Cross-validation yielded high classification accuracies (F1-score>0.95) for both the k-NN and SVM classifiers in detecting compensation movements during all the reaching tasks. In detail, an excellent performance was achieved in discriminating between compensation and noncompensation (NC) movements, with an average F1score of 0.993. For the multiclass classification of compensatory movement patterns, an average F1-score of 0.981 was achieved in recognizing the NC, trunk lean-forward (TLF), trunk rotation (TR) and shoulder elevation (SE) movements. Conclusions: Good classification performance in detecting and categorizing compensatory movements validated the feasibility of the proposed pressure distribution-based system. Reliable classification accuracy achieved by the machine learning algorithms indicated the potential to monitor compensation movements automatically by using the pressure distribution-based system when stroke survivors perform seated reaching tasks.

show abstract

Section: Introductionmentioning

confidence: 99%

Detecting compensatory movements of stroke survivors using pressure distribution data and machine learning algorithms

Cai

Zhang

et al. 2019

J NeuroEngineering Rehabil

View full text Add to dashboard Cite

show abstract

“…For this, we plan to extend our usability study in which we will acquire electromyogram signals from one's upper extremity during execution of such tasks. Thus, added to our preliminary calibration of one's individualized range of motion with the horizontal span of the HCI Task monitor, we plan to perform biomechanical evaluation of the patients similar to that in other studies (Kim et al, 2016; Bakhti et al, 2018; Scano et al, 2018) to tune individualized thresholds. Again, with regard to study settings, our present study offered a limited number of exposures to the post-stroke participants.…”

Section: Discussionmentioning

confidence: 99%

Kinect-Assisted Performance-Sensitive Upper Limb Exercise Platform for Post-stroke Survivors

et al. 2019

View full text Add to dashboard Cite

One's ability to use upper limbs is critical for performing activities of daily living necessary for enjoying quality community life. However, after stroke, such abilities becomes adversely affected and it often deprives one of their capability to perform tasks that need coordinated movement in the upper limbs. To address issues with upper limb dysfunction, patients typically undergo rehabilitative exercises. Given the high patient to doctor ratio particularly in developing countries like India, conventional rehabilitation with patients undergoing exercises under one-on-one therapist's supervision often becomes a challenge. Thus, investigators are exploring technology such as computer-based platforms coupled with cameras that can alleviate the need for the continuous presence of a therapist and can offer a powerful complementary tool in the hands of the clinicians. Such marker-based imaging systems used for rehabilitation can offer real-time processing and high accuracy of data. However, these systems often require dedicated lab space and high set-up time. Often this is very expensive and suffers from portability issues. Investigators have been exploring marker-less imaging techniques e.g., Kinect integrated computer-based graphical user interfaces in stroke-rehabilitation such as tracking one's limb movement during rehabilitation. In our present study, we have developed a Kinect-assisted computer-based system that offered Human Computer Interaction (HCI) tasks of varying challenge levels. Execution of the tasks required one to use reaching and coordination skills of the upper limbs. Also, the system was Performance-sensitive i.e., adaptive to the individualized residual movement ability of one's upper limb quantified in terms of task performance score. We tested for the usability of our system by exposing 15 healthy participants to our system. Subsequently, seven post-stroke patients interacted with our system over a few sessions spread over 2 weeks. Also, we studied patient's mean tonic activity corresponding to the HCI tasks as a possible indicator of one's post-stroke functional recovery suggesting its potential of our system to serve as a rehabilitation platform. Our results indicate the potential of such systems toward the improvement of task performance capability of post-stroke patients with possibilities of upper limb movement rehabilitation.

show abstract

“…The results of this pilot study showed that our device using Kinect v2 accurately measured ataxic movements employed in SARA ( Table 1) with high accuracy of better than 2 mm by simple comparison with a ruler. On the other hand, previous studies that validated accuracy of Kinect used other motion captures such as CMS20s (Zebris, Germany) for forelimb movement of stroke patients (19) or Optotrak Certus System (Northern Digital, Canada) for gait of healthy controls (21). However, they were not able to provide absolute accuracy in the measurement of Kinect.…”

Section: Reproducibility and Multiplicity In Digitalized Saramentioning

confidence: 99%

Assessment and Rating of Motor Cerebellar Ataxias With the Kinect v2 Depth Sensor: Extending Our Appraisal

et al. 2020

View full text Add to dashboard Cite

show abstract

Kinect-based assessment of proximal arm non-use after a stroke

Cited by 48 publications

References 64 publications

Detecting compensatory movements of stroke survivors using pressure distribution data and machine learning algorithms

Detecting compensatory movements of stroke survivors using pressure distribution data and machine learning algorithms

Kinect-Assisted Performance-Sensitive Upper Limb Exercise Platform for Post-stroke Survivors

Assessment and Rating of Motor Cerebellar Ataxias With the Kinect v2 Depth Sensor: Extending Our Appraisal

Contact Info

Product

Resources

About