Quantification of abnormal upper limb movement during walking in people with acquired brain injury

Machine learning for automating subjective clinical assessment of gait impairment in people with acquired brain injury - a comparison of an image extraction and classification system to expert scoring

BACKGROUND

Walking impairment is a common disability post acquired brain injury (ABI), with visually evident arm movement abnormality identified as negatively impacting a multitude of psychological factors. The International Classification of Functioning, Disability and Health (ICF) qualifiers scale has been used to subjectively assess arm movement abnormality, showing strong intra-rater and test-retest reliability, however, only moderate inter-rater reliability. This impacts clinical utility, limiting its use as a measurement tool. To both automate the analysis and overcome these errors, the primary aim of this study was to evaluate the ability of a novel two-level machine learning model to assess arm movement abnormality during walking in people with ABI.

METHODS

Frontal plane gait videos were used to train four networks with 50%, 75%, 90%, and 100% of participants (ABI: n = 42, healthy controls: n = 34) to automatically identify anatomical landmarks using DeepLabCut™ and calculate two-dimensional kinematic joint angles. Assessment scores from three experienced neurorehabilitation clinicians were used with these joint angles to train random forest networks with nested cross-validation to predict assessor scores for all videos. Agreement between unseen participant (i.e. test group participants that were not used to train the model) predictions and each individual assessor's scores were compared using quadratic weighted kappa. One sample t-tests (to determine over/underprediction against clinician ratings) and one-way ANOVA (to determine differences between networks) were applied to the four networks.

RESULTS

The machine learning predictions have similar agreement to experienced human assessors, with no statistically significant (p < 0.05) difference for any match contingency. There was no statistically significant difference between the predictions from the four networks (F = 0.119; p = 0.949). The four networks did however under-predict scores with small effect sizes (p range = 0.007 to 0.040; Cohen's d range = 0.156 to 0.217).

CONCLUSIONS

This study demonstrated that machine learning can perform similarly to experienced clinicians when subjectively assessing arm movement abnormality in people with ABI. The relatively small sample size may have resulted in under-prediction of some scores, albeit with small effect sizes. Studies with larger sample sizes that objectively and automatically assess dynamic movement in both local and telerehabilitation assessments, for example using smartphones and edge-based machine learning, to reduce measurement error and healthcare access inequality are needed.

Comparison of wrist actimetry variables of paretic upper limb use in post stroke patients for ecological monitoring

BACKGROUND

To date, many wrist actimetric variables dedicated to measuring the upper limbs (UL) in post-stroke patients have been developed but very few comparisons have been made between them. The objective of this study was to compare different actimetric variables of the ULs between a stroke and healthy population.

METHODS

Accelerometers were worn continuously for a period of 7 days on both wrists of 19 post-stroke hemiparetic patients as well as 11 healthy subjects. Various wrist actimetry variables were calculated, including the Jerk ratio 50 (JR50, cumulative probability that the Jerk Ratio is between 1 and 2), absolute (FuncUse30) and relative (FuncUseRatio30) amounts of functional use of movements of the ULs with angular amplitude greater than 30°, and absolute (UH) and relative (UseHoursRatio) use hours.

RESULTS

FuncUse30, FuncUseRatio30, UH, UseHoursRatio and JR50 of the paretic UL of stroke patients were significantly lower than in the non-dominant UL of healthy subjects. Comparing the ratio variables in stroke patients, FuncUseRatio30 was significantly lower than UseHoursRatio and JR50, suggesting a more clinically sensitive variable to monitor. In an exploratory analysis, FuncUseRatio tends to decrease with angular range of motion for stroke patients while it remains stable and close to 1 for healthy subjects. UseHoursRatio, FuncUseRatio30 and JR50 show linear correlation with Fugl-Meyer score (FM), with r² equal to 0.53, 0.35 and 0.21, respectively.

CONCLUSION

This study determined that the FuncUseRatio30 variable provides the most sensitive clinical biomarker of paretic UL use in post-stroke patients, and that FuncUseHours-angular range of motion relationship allows the identification of the UL behaviour of each patient. This ecological information on the level of functional use of the paretic UL can be used to improve follow-up and develop patient-specific therapy.

Considerations for ensuring safety during telerehabilitation of people with stroke. A protocol for a scoping review

INTRODUCTION

Exercise interventions have a positive impact on people with stroke. However, access to exercise interventions is variable, and there may be a delay in the start of rehabilitation. Telerehabilitation has enabled the delivery of exercise interventions replacing the traditional face-to-face approach. Aspects related to the safety of people with stroke should be considered to avoid adverse events during the delivery of exercise interventions remotely. However, such information is scattered in the literature, and the detail with which measures taken during the implementation of exercise interventions for people with stroke are reported is unknown.

OBJECTIVE

To summarise measures or aspects targeted at reducing the incidence of adverse events during the delivery of exercise interventions through telerehabilitation in patients after stroke.

MATERIALS AND METHODS

A scoping review will be conducted. A systematic search in MEDLINE-Ovid, Embase-Ovid CENTRAL, CINAHL Complete (EBSCOhost), and other resources will be carried out. We will include primary studies, published in full text in any language, involving people with stroke who undergo telerehabilitation where exercise is the main component. Two reviewers will independently select studies and extract data, and disagreements will be resolved by consensus or a third reviewer. The results will be reported in a narrative form, using tables and figures to support them.

DISCUSSION

To implement this strategy within rehabilitation services, one of the first aspects to be solved is to ensure the safety of people. The results of this scoping review could contribute an information base for clinicians and decision-makers when designing remotely delivered exercise intervention programs.

REGISTRATION NUMBER

INPLASY202290104.

Treatment of spasticity

Spasticity is characterized by an enhanced size and reduced threshold for activation of stretch reflexes and is associated with "positive signs" such as clonus and spasms, as well as "negative features" such as paresis and a loss of automatic postural responses. Spasticity develops over time after a lesion and can be associated with reduced speed of movement, cocontraction, abnormal synergies, and pain. Spasticity is caused by a combination of damage to descending tracts, reductions in inhibitory activity within spinal cord circuits, and adaptive changes within motoneurons. Increased tone, hypertonia, can also be caused by changes in passive stiffness due to, for example, increase in connective tissue and reduction in muscle fascicle length. Understanding the cause of hypertonia is important for determining the management strategy as nonneural, passive causes of stiffness will be more amenable to physical rather than pharmacological interventions. The management of spasticity is determined by the views and goals of the patient, family, and carers, which should be integral to the multidisciplinary assessment. An assessment, and treatment, of trigger factors such as infection and skin breakdown should be made especially in people with a recent change in tone. The choice of management strategies for an individual will vary depending on the severity of spasticity, the distribution of spasticity (i.e., whether it affects multiple muscle groups or is more prominent in one or two groups), the type of lesion, and the potential for recovery. Management options include physical therapy, oral agents; focal therapies such as botulinum injections; and peripheral nerve blocks. Intrathecal baclofen can lead to a reduction in required oral antispasticity medications. When spasticity is severe intrathecal phenol may be an option. Surgical interventions, largely used in the pediatric population, include muscle transfers and lengthening and selective dorsal root rhizotomy.

Upper Limb Associated Reactions: The Relationship Between Movement Kinematics and Muscle Activity in Seated Versus Walking Testing

OBJECTIVE

The aims of this study were to (1) evaluate the relationships between stationary and dynamic associated reaction (AR) tests in people with acquired brain injury using surface electromyography (sEMG) muscle activity and three-dimensional motion analysis kinematic measures and (2) assess the test-retest reliability of sEMG and seated tests of ARs.

DESIGN

Forty-two adults with acquired brain injury underwent AR testing with seated contralateral maximal voluntary isometric contraction tests and walking (self-selected and fast speeds). Associated reaction measurements included biceps brachii sEMG, elbow goniometry, and three-dimensional motion analysis kinematics during walking. Pearson correlations evaluated the relationships between seated and dynamic walking AR tests and between muscle activity and kinematic measures. Chronic participants were reassessed 1 wk later for reliability.

RESULTS

A strong (r = 0.65) and moderate (r = 0.53) relationship existed for biceps brachii sEMG during seated and walking tests at self-selected and fast walk, respectively. A weak to moderate relationship existed between biceps brachii sEMG and kinematics during walking and between seated and walking measures of ARs (r = 0.23-0.53). All tests had strong to very strong test-retest reliability (intraclass correlation coefficients, >0.78).

CONCLUSION

Seated contralateral maximal voluntary isometric contraction tests correlate only weakly to moderately with AR walking kinematics and moderately to strongly with biceps brachii activation during walking. Moderate relationships exist between sEMG and kinematics, indicating that they may provide different information for ARs.

Potential contributing factors to upper limb associated reactions in people with acquired brain injury: an exploratory study

PURPOSE

To determine which potential contributing factors are associated with upper limb associated reaction (AR) expression in individuals with acquired brain injury (ABI).

METHODS

Forty-two participants underwent three-dimensional motion analysis at self-selected walking speed to generate the AR outcome measure, quantifying their upper limb kinematic deviation compared to healthy controls. Clinical assessment included: upper and lower limb hypertonicity, spasticity and strength, balance, dynamic walking stability, arm and leg function, anxiety, arm pain/discomfort, and fear of falling.

RESULTS

Significant, moderate-to-strong correlations (r = 0.42-0.74, p < 0.05) existed between upper limb ARs and both hypertonicity and spasticity of the upper limb muscles and the knee extensors. Significant, moderate correlations to ARs (r = 0.42-0.59, p < 0.05) existed for balance, dynamic stability, upper limb strength, and arm function. The severity of AR was significantly different between those with and without hypertonicity of the four tested upper limb muscles, elbow and long finger flexor spasticity, knee extensor spasticity, and reduced dynamic stability (p < 0.05; effect sizes ≥0.80). However, these contributing factors were not present in all participants.

CONCLUSIONS

Associated reactions are complex and multi-factorial. There were several significant correlations indicating that factors may influence AR severity. While positive upper motor neuron syndrome features should be prioritised for clinical assessment, these factors are not prerequisites for ARs.IMPLICATIONS FOR REHABILITATIONUpper limb associated reactions are a complex and multi-factorial phenomenon.Upper limb muscle hypertonicity and spasticity should be prioritised for assessment; however, they are not prerequisites for associated reactions.Hypertonicity and spasticity should be differentiated as they may have differing relationships to associated reactions.Knee extensor hypertonicity and spasticity, postural stability, upper limb strength, and arm function may also be contributing factors to consider.