Comparison of Machine Learning approaches for Classifying Upper Extremity Tasks in Individuals Post-Stroke

Classification of functional and non-functional arm use by inertial measurement units in individuals with upper limb impairment after stroke

Background: Arm use metrics derived from wrist-mounted movement sensors are widely used to quantify the upper limb performance in real-life conditions of individuals with stroke throughout motor recovery. The calculation of real-world use metrics, such as arm use duration and laterality preferences, relies on accurately identifying functional movements. Hence, classifying upper limb activity into functional and non-functional classes is paramount. Acceleration thresholds are conventionally used to distinguish these classes. However, these methods are challenged by the high inter and intra-individual variability of movement patterns. In this study, we developed and validated a machine learning classifier for this task and compared it to methods using conventional and optimal thresholds. Methods: Individuals after stroke were video-recorded in their home environment performing semi-naturalistic daily tasks while wearing wrist-mounted inertial measurement units. Data were labeled frame-by-frame following the Taxonomy of Functional Upper Limb Motion definitions, excluding whole-body movements, and sequenced into 1-s epochs. Actigraph counts were computed, and an optimal threshold for functional movement was determined by receiver operating characteristic curve analyses on group and individual levels. A logistic regression classifier was trained on the same labels using time and frequency domain features. Performance measures were compared between all classification methods. Results: Video data (6.5 h) of 14 individuals with mild-to-severe upper limb impairment were labeled. Optimal activity count thresholds were ≥20.1 for the affected side and ≥38.6 for the unaffected side and showed high predictive power with an area under the curve (95% CI) of 0.88 (0.87,0.89) and 0.86 (0.85, 0.87), respectively. A classification accuracy of around 80% was equivalent to the optimal threshold and machine learning methods and outperformed the conventional threshold by ∼10%. Optimal thresholds and machine learning methods showed superior specificity (75-82%) to conventional thresholds (58-66%) across unilateral and bilateral activities. Conclusion: This work compares the validity of methods classifying stroke survivors' real-life arm activities measured by wrist-worn sensors excluding whole-body movements. The determined optimal thresholds and machine learning classifiers achieved an equivalent accuracy and higher specificity than conventional thresholds. Our open-sourced classifier or optimal thresholds should be used to specify the intensity and duration of arm use.

Classifying Unimanual and Bimanual Upper Extremity Tasks in Individuals Post-Stroke

After stroke, many individuals develop impairments that lead to compensatory motions. Compensation allows individuals to achieve tasks but has long-term detrimental effects and represents maladaptive motor strategies. Increased use of bimanual motions may serve as a biomarker for recovery (and the reduction of reliance on compensatory motion), and tracking such motion using sensor data may provide critical data for health care specialists. However, past work by the authors demonstrated individual variation in motor strategies results in noisy and chaotic sensor data. The goal of the current work is to develop classifiers capable of differentiating unimanual, bimanaual asymmetric, and bimanual symmetric gestures using wearable sensor data. Twenty participants post-stroke (and 20 age-matched controls) performed a set of tasks under the supervision of a trained occupational therapist. Sensor data were recorded for each task. Classifiers were developed using artificial neural networks (ANNs) as a baseline, and the echo state neural network (ESNN) which has demonstrated efficacy with chaotic data. We find that, for control and post-stroke participants, the ESNN results in improved testing accuracy performance (91.3% and 80.3%, respectively). These results suggest a novel method for classifying gestures in individuals post-stroke, and the developed classifiers may facilitate longitudinal monitoring and correction of compensatory motion.