Parameters that remain consistent independent of pausing before gait-initiation during normal rise-to-walk behaviour delineated by sit-to-walk and sit-to-stand-and-walk

Sit-to-walk strategy classification in healthy adults using hip and knee joint angles at gait initiation

Forward continuation, balance, and sit-to-stand-and-walk (STSW) are three common movement strategies during sit-to-walk (STW) executions. Literature identifies these strategies through biomechanical parameters using gold standard laboratory equipment, which is expensive, bulky, and requires significant post-processing. STW strategy becomes apparent at gait-initiation (GI) and the hip/knee are primary contributors in STW, therefore, this study proposes to use the hip/knee joint angles at GI as an alternate method of strategy classification. To achieve this, K-means clustering was implemented using three clusters corresponding to the three STW strategies; and two feature sets corresponding to the hip/knee angles (derived from motion capture data); from an open access online database (age: 21-80 years; n = 10). The results identified forward continuation with the lowest hip/knee extension, followed by balance and then STSW, at GI. Using this classification, strategy biomechanics were investigated by deriving the established biomechanical quantities from literature. The biomechanical parameters that significantly varied between strategies (P < 0.05) were time, horizontal centre of mass (COM) momentum, braking impulse, centre of pressure (COP) range and velocities, COP-COM separation, hip/knee torque and movement fluency. This alternate method of strategy classification forms a generalized framework for describing STW executions and is consistent with literature, thus validating the joint angle classification method.

Single Accelerometer to Recognize Human Activities Using Neural Networks

Exoskeletons have decreased physical effort and increased comfort in activities of daily living (ADL) such as walking, squatting, and running. However, this assistance is often activity specific and does not accommodate a wide variety of different activities. To overcome this limitation and increase the scope of exoskeleton application, an automatic human activity recognition (HAR) system is necessary. We developed two deep-learning models for HAR using one-dimensional-convolutional neural network (CNN) and a hybrid model using CNNs and long-short term memory (LSTM). We trained both models using the data collected from a single three-axis accelerometer placed on the chest of ten subjects. We were able to classify five different activities, standing, walking on level ground, walking on an incline, running, and squatting, with an accuracy of 98.1% and 97.8%, respectively. A two subject real-time validation trial was also conducted to validate the real-time applicability of the system. The real-time accuracy was measured at 96.6% and 97.2% for the CNN and the hybrid model, respectively. The high classification accuracy in the test and real-time evaluation suggests that a single sensor could distinguish human activities using machine-learning-based models.

Investigating the Relationships Between Three Important Functional Tasks Early After Stroke: Movement Characteristics of Sit-To-Stand, Sit-To-Walk, and Walking

Background: Walking, sit-to-stand (STS) and sit-to-walk (STW) are all considered important functional tasks in achieving independence after stroke. Despite knowledge that sensitive measurement of movement patterns is crucial to understanding neuromuscular restitution, there is surprisingly little information available about the detailed biomechanical characteristics of, and relationships between, walking, sit-to-stand and sit-to-walk, particularly in the important time window early after stroke. Hence, here, the study aimed to:

Identify the biomechanical characteristics of and determine any differences in both movement fluidity (hesitation, coordination and smoothness) and duration of movement phases, between sit-to-stand (STS) and sit-to-walk (STW) in people early after stroke.

Determine whether measures of movement fluidity (hesitation, coordination, and smoothness) and movement phases during sit-to-stand (STS) and/or sit-to-walk (STW) are correlated strongly to commonly used measures of walking speed and/or step length ratio in people early after stroke.

Methods: This study consisted of secondary data analysis from the SWIFT Cast Trial. Specifically, we investigated movement fluidity using established assessments of smoothness, hesitation and coordination and the time duration for specific movement phases in a group of 48 people after stroke. Comparisons were made between STS and STW and relationships to walking measures were explored.

Results: Participants spent significantly more time in the initial movement phase, flexion momentum, during STS [mean time (SD) 1.74 ±1.45 s] than they did during STW [mean time (SD) 1.13 ± 1.03 s]. STS was also completed more smoothly but with more hesitation and greater coordination than the task of STW. No strong relationships were found between movement fluidity or duration with walking speed or step length symmetry.

Conclusions: Assessment of movement after stroke requires a range of functional tasks and no one task should predominate over another. Seemingly similar or overlapping tasks such as STS and STW create distinct biomechanical characteristics which can be identified using sensitive, objective measures of fluidity and movement phases but there are no strong relationships between the functional tasks of STS and STW with walking speed or with step-length symmetry.