Evaluation of the validity, reliability, and feasibility of an inertial sensor system for trunk assessment and rehabilitation post-stroke

Abstract

Post stroke trunk control is recognised as an important independent predictor of activities of daily living after a stroke. Clinical scales such as the Trunk Impairment Scale (TIS) are recommended to be used in clinical research to assess post stroke trunk impairment; however, they do not consider the quality of trunk movement. Assessing trunk and pelvic movement quality would inform our understanding of trunk recovery and, therefore, facilitate the development of more effective rehabilitation interventions. Three-dimensional kinematic measurement systems are used to measure trunk movement quality in research laboratories, however, these are not suitable for implementation in clinical settings due to the high cost and physical space requirements. The use of inertial sensors to measure trunk performance have the potential to overcome these limitations and, therefore, offer an alternative solution to measure trunk performance. The aim of this research is to investigate the use of the Valedo® system to assess trunk performance and, when combined with video games, as an intervention tool to deliver trunk exercises for people with chronic stroke.In the first stage of this PhD study, a cross-sectional validity and reliability study investigated the development of an instrumented TIS (iTIS) and, subsequently, tested its validity and reliability in 20 patients with chronic stroke and 20 aged-matched healthy participants. The kinematic parameters of the trunk and the clinical TIS (cTIS) were measured simultaneously. A moderate relationship was found between the observed iTIS parameters and the clinical TIS scores, supporting the concurrent validity of the iTIS. The reliability for the dynamic subscale parameters was good to excellent (intrarater ICC = 0.60–0.95; interrater ICC = 0.59–0.93); however, reliability for the coordination parameters was poor to good (intrarater ICC = 0.05–0.72) and poor to excellent (interrater ICC = 0.04–0.78). These results demonstrate that the iTIS has important implications for objective assessments of trunk impairment in clinical practice; the iTIS can be used in conjunction with the clinical TIS to detect small changes in trunk range of movement that cannot be observed clinically.In the second stage of this PhD study, a cross-sectional validity and reliability study investigated the use of Valedo® sensors to measure trunk movement during streamlined Wolf Motor Function Test (SWMFT) performance. A sample of 20 people with chronic stroke and 20 age-matched healthy participants performed the SWMFT while wearing Valedo® sensors on their trunk to capture trunk movements. The Valedo® system was able to distinguish between the stroke participants and the healthy participants; the stroke participants employed greater trunk range of movements than the healthy controls in all tasks (p < 0.01). Furthermore, using trunk parameters recorded by the Valedo® system enabled differentiation between the affected hand and the unaffected hand of people within the stroke group. The reliability for the stroke group was good to excellent with intrarater reliability (same session; ICC = 0.70–0.95): intrarater reliability (between two days; ICC = 0.71–0.92) and interrater reliability (between two raters; ICC = 0.63–0.95). These findings indicate that the Valedo system has important potential for the objective assessment of trunk range of movement during upper limb (UL) activity in clinical practice.In the third stage of this PhD study, a convergent parallel mixed method study investigated the feasibility of conducting a programme of 18 sessions of trunk exercises using a video game–based system (Valedo) for people with stroke. It also explored possible changes in clinical outcome measures involving trunk impairment (TIS and iTIS), trunk muscle size, balance and lower limb (Fugl-Meyer Assessment of the lower limb) and upper limb function, as well as impairment outcomes (SWMFT and Fugl-Meyer Assessment of the upper limb). Of the 10 chronic stroke survivors who participated in the intervention, 8 completed 18 sessions. Post intervention results showed adherence was excellent, no major adverse events occurred and very good-to-excellent participation (the Pittsburgh Rehabilitation Participation Scale) was achieved. The clinical outcome measures demonstrated an improvement in trunk impairment (TIS and iTIS), balance (the Berg Balance Scale) and upper limb function (SWMFT). The post intervention interview revealed a high acceptability of the intervention, associated with enjoyment and perceived physical and psychological benefits from the intervention. The use of audio-visual feedback was perceived as helpful to set a goal and motivated participants to engage in the exercises. Several points need to be considered when implementing this type of intervention in the future. For the person administering the intervention, it includes personalising the task difficulty level appropriately to avoid excessive physical and cognitive challenges and ensuring that participants understand the system-generated feedback on performance following each game. For game designers, game reward strategies need to be improved to minimise participants’ frustration. In addition, using artificial intelligence in the system will help with tailoring the exercise programmes by suggesting to the therapists the most appropriate games to practise next.The three stages of this PhD study addressed the original aim of this research and provided evidence that the Valedo® system can be used to assess trunk performance and, when combined with video games, and feasible to use as an intervention tool to deliver trunk exercises using video games for people with chronic stroke.

More information

Identifiers

ORCID for Ann-Marie Hughes: orcid.org/0000-0002-3958-8206

Catalogue record

Export record