Dual-Tasking in Daily Activities Among Adults With and Without Stroke.

IMPORTANCE: In laboratory settings, dual-tasking is a performance strategy affected by dominance and stroke. However, the volitional use of dual-tasking has not been examined during naturalistic performance of activities of daily living (ADLs). OBJECTIVE: To examine dual-tasking in the context of ADLs and identify whether dominance and stroke influence its use. DESIGN: Cross-sectional, observational. SETTING: Academic medical center. PARTICIPANTS: Forty-three participants with chronic stroke and upper extremity (UE) motor impairment and 19 control participants without stroke. OUTCOMES AND MEASURES: We identified dual-tasking as the performance of dual-object primitives (DOPs), a functional strategy to manage two objects simultaneously. We videotaped participants performing feeding and toothbrushing tasks and identified the initiation and frequency of DOPs. We assessed whether these outcomes were influenced by UE dominance or paresis and whether among participants with stroke these outcomes were influenced by motor impairment (using the Fugl-Meyer Assessment) or cognitive impairment (using the Montreal Cognitive Assessment). RESULTS: DOP initiation was reduced on the nondominant side of control UEs and in the paretic UE of participants with stroke. After DOPs were initiated, however, their frequency was not significantly related to dominance or paresis. Among participants with stroke, DOP initiation but not DOP frequency was influenced by motor impairment, and neither were influenced by cognitive impairment. CONCLUSIONS AND RELEVANCE: The initiation of dual-tasking is curtailed in the nondominant and paretic UEs, extending previous laboratory-based findings to a more naturalistic setting. These results may reflect a demand on neural resources that is exceeded when these limbs are used. What This Article Adds: DOPs, a functional strategy to simultaneously engage two objects during ADLs, could serve as a behavioral marker of dual-tasking in real-world activities, supporting their investigation more broadly. Practicing DOPs in rehabilitation could also train the integration of dual-tasking strategies in activity execution.

PrimSeq: A deep learning-based pipeline to quantitate rehabilitation training.

Stroke rehabilitation seeks to accelerate motor recovery by training functional activities, but may have minimal impact because of insufficient training doses. In animals, training hundreds of functional motions in the first weeks after stroke can substantially boost upper extremity recovery. The optimal quantity of functional motions to boost recovery in humans is currently unknown, however, because no practical tools exist to measure them during rehabilitation training. Here, we present PrimSeq, a pipeline to classify and count functional motions trained in stroke rehabilitation. Our approach integrates wearable sensors to capture upper-body motion, a deep learning model to predict motion sequences, and an algorithm to tally motions. The trained model accurately decomposes rehabilitation activities into elemental functional motions, outperforming competitive machine learning methods. PrimSeq furthermore quantifies these motions at a fraction of the time and labor costs of human experts. We demonstrate the capabilities of PrimSeq in previously unseen stroke patients with a range of upper extremity motor impairment. We expect that our methodological advances will support the rigorous measurement required for quantitative dosing trials in stroke rehabilitation.

StrokeRehab: A Benchmark Dataset for Sub-second Action Identification.

Automatic action identification from video and kinematic data is an important machine learning problem with applications ranging from robotics to smart health. Most existing works focus on identifying coarse actions such as running, climbing, or cutting vegetables, which have relatively long durations and a complex series of motions. This is an important limitation for applications that require identification of more elemental motions at high temporal resolution. For example, in the rehabilitation of arm impairment after stroke, quantifying the training dose (number of repetitions) requires differentiating motions with sub-second durations. Our goal is to bridge this gap. To this end, we introduce a large-scale, multimodal dataset, StrokeRehab, as a new action-recognition benchmark that includes elemental short-duration actions labeled at a high temporal resolution. StrokeRehab consists of high-quality inertial measurement unit sensor and video data of 51 stroke-impaired patients and 20 healthy subjects performing activities of daily living like feeding, brushing teeth, etc. Because it contains data from both healthy and impaired individuals, StrokeRehab can be used to study the influence of distribution shift in action-recognition tasks. When evaluated on StrokeRehab, current state-of-the-art models for action segmentation produce noisy predictions, which reduces their accuracy in identifying the corresponding sequence of actions. To address this, we propose a novel approach for high-resolution action identification, inspired by speech-recognition techniques, which is based on a sequence-to-sequence model that directly predicts the sequence of actions. This approach outperforms current state-of-the-art methods on StrokeRehab, as well as on the standard benchmark datasets 50Salads, Breakfast, and Jigsaws.

A Taxonomy of Functional Upper Extremity Motion.

Background: Functional upper extremity (UE) motion enables humans to execute activities of daily living (ADLs). There currently exists no universal language to systematically characterize this type of motion or its fundamental building blocks, called functional primitives. Without a standardized classification approach, pooling mechanistic knowledge and unpacking rehabilitation content will remain challenging. Methods: We created a taxonomy to characterize functional UE motions occurring during ADLs, classifying them by motion presence, temporal cyclicity, upper body effector, and contact type. We identified five functional primitives by their phenotype and purpose: reach, reposition, transport, stabilize, and idle. The taxonomy was assessed for its validity and interrater reliability in right-paretic chronic stroke patients performing a selection of ADL tasks. We applied the taxonomy to identify the primitive content and motion characteristics of these tasks, and to evaluate the influence of impairment level on these outcomes. Results: The taxonomy could account for all motions in the sampled activities. Interrater reliability was high for primitive identification (Cohen's kappa = 0.95-0.99). Using the taxonomy, the ADL tasks were found to be composed primarily of transport and stabilize primitives mainly executed with discrete, proximal motions. Compared to mildly impaired patients, moderately impaired patients used more repeated reaches and axial-proximal UE motion to execute the tasks. Conclusions: The proposed taxonomy yields objective, quantitative data on human functional UE motion. This new method could facilitate the decomposition and quantification of UE rehabilitation, the characterization of functional abnormality after stroke, and the mechanistic examination of shared behavior in motor studies.