Advanced Techniques for Robotic Assessment of Neurological Impairments In Stroke Patients

Thumbnail Image

Authors

Tyryshkin, Kathrin

Date

2014-09-29

Type

thesis

Language

eng

Keyword

task analysis , data mining , fuzzy neural network , feature selection , stroke assessment , task comparison , feature information , robotic technology

Research Projects

Organizational Units

Journal Issue

Alternative Title

Abstract

Stroke is an acute injury of the central nervous system and is caused by the disruption of blood flow or by the rupture of blood vessels. A stroke can impact many body functions, often causing motor, speech, memory, vision and other sensory impairments. It is highly prevalent and often requires long hospitalizations and post-stroke rehabilitation. The key to successful rehabilitation after stroke is an accurate assessment of the impairment. Current clinical assessments of stroke-related impairments involve physical assessment and visual observation by physicians. Existing clinical scores of upper limb function often use observer-based ordinal scales that are subjective and commonly have floor and ceiling effects. Therefore, these methods are not adequate to reliably discriminate different levels of performance. Robotics and integrated virtual reality systems have a tremendous potential to be used in computational systems that analyze, visualize and aid clinicians to identify and assess sensory-motor impairments. This thesis presents a framework for analysis and extraction of reliable and valid features from robotic data that can be used to accurately and objectively assess neurological impairments. The framework was applied on the Object Hit task that assesses the ability of participants to select and engage motor actions with both hands. In addition, the Object Hit Task was compared to the Object Hit and Avoid task, which is a slight modification of the original task. The comparison was done using a developed feature and task analysis framework. This framework encompasses similarities and differences between tasks in a given experiment in terms of feature information. The results showed that for the data used in the analysis, Object Hit task is able to identify impairments more effectively than the Object Hit and Avoid task. The overall results demonstrate that the Object Hit task provides an objective and easy approach to quantify upper limb motor function and visuospatial skills after stroke. The developed assessment tool can also be applied for diagnosis and prognosis of other neurological deficits, beyond stroke assessment.

Description

Thesis (Ph.D, Computing) -- Queen's University, 2014-09-29 17:06:09.217

Citation

Publisher

License

This publication is made available by the authority of the copyright owner solely for the purpose of private study and research and may not be copied or reproduced except as permitted by the copyright laws without written authority from the copyright owner.

Journal

Volume

Issue

PubMed ID

External DOI

ISSN

EISSN