PaperlayerView exact revision on arXiv

Source-linked AI summary

Lower-limb kinematics and kinetics during continuously varying human locomotion

Emma Reznick, Kyle R. Embry, Ross Neuman, Edgar Bolívar-Nieto, Nicholas P. Fey, Robert D. Gregg

arXiv:2108.12307v1cs.RO

TL;DR

Existing locomotion datasets often omit task transitions or continuous speed and incline variation, although these conditions are relevant to prosthetic and exoskeleton development. This paper presents lower-limb measurements from ten able-bodied participants across varied locomotion tasks and transitions. The dataset includes kinematic and kinetic measures where available and is intended to support continuous locomotion modeling.

  • Problem

    Available locomotion datasets omit task transitions and/or continuous speed and incline variation, despite their relevance to developing robotic prostheses and exoskeletons.

    Source: Background & Summary · pp. 1–2

  • Method

    The authors recorded lower-limb motion from ten able-bodied participants across varied walking, running, stair, and transition tasks, using motion capture and treadmill force plates where applicable.

    Source: Background & Summary · p. 2, Motion Capture · p. 3

  • Results

    The resulting dataset spans sit-stand, walk-run, and walk-stairs transitions, multiple walking speeds and inclines, running speeds, and stair inclines.

    Source: Background & Summary · p. 2, Walking · p. 3, Running · p. 4

  • Takeaways & Limitations

    The dataset can serve as input for models that interpolate walking kinematics across continuously varying speeds and inclines and model transitions between locomotion tasks.

    Source: Background & Summary · p. 2

    The dataset comprises ten able-bodied participants rather than patient populations.

    Source: Background & Summary · p. 2

Abstract

from arXiv · show

Human locomotion involves continuously variable activities including walking, running, and stair climbing over a range of speeds and inclinations as well as sit-stand, walk-run, and walk-stairs transitions. Understanding the kinematics and kinetics of the lower limbs during continuously varying locomotion is fundamental to developing robotic prostheses and exoskeletons that assist in community ambulation. However, available datasets on human locomotion neglect transitions between activities and/or continuous variations in speed and inclination during these activities. This data paper reports a new dataset that includes the lower-limb kinematics and kinetics of ten able-bodied participants walking at multiple inclines ($\pm$ 0, 5, 10 $^{\circ}$) and speeds (0.8, 1, 1.2 m/s), running at multiple speeds (1.8, 2, 2.2, 2.4 m/s), walking and running with constant acceleration ($\pm$ 0.2, 0.5 $\text{m/s^2}$), and stair ascent/descent with multiple stair inclines (20, 25, 30, 35 $^{\circ}$). This dataset also includes sit-stand transitions, walk-run transitions, and walk-stairs transitions. Data were recorded by a Vicon motion capture system and, for applicable tasks, a Bertec instrumented treadmill.

Background & Summary

Existing locomotion datasets often omit transitions and continuous changes in speed or incline, despite their relevance to modeling everyday movement and prosthetic control. This paper presents a ten-person able-bodied dataset spanning varied locomotion tasks and transitions, with measurements intended to support continuous kinematic modeling.

Background & Summary

  • Non-steady locomotion includes intermittent bouts of walking, stopping, sitting, standing, and stair climbing; 75% of walking bouts are fewer than 40 steps.The paper identifies transitions and continuous speed and slope changes as relevant to modeling locomotion and designing agile prostheses.Source: Background & Summary · p. 1
  • Existing datasets omit one or more of joint kinetics, global orientation, older participants, activities of daily living, task transitions, or continuous speed and incline variation.The paper notes that datasets are difficult to combine because they differ in measurements, methods, and participants.Source: Background & Summary · pp. 1–2
  • The dataset records lower-limb measurements from ten able-bodied participants across sit-stand, walk-run, and walk-stairs transitions, varied walking speeds and inclines, and running speeds and accelerations.It includes pelvic tilt, hip, knee, and ankle angles, joint moments for treadmill tasks, 3D marker positions, and force-plate measurements when available.Source: Background & Summary · p. 2
  • Continuous models of gait can interpolate walking kinematics across varying speeds and inclines, unlike control based on switching among a limited set of discrete conditions.The paper describes using participant-specific tasks alongside across-participant averages to individualize kinematic models for prostheses and exoskeletons.Source: Background & Summary · p. 2

Methods

The study acquired motion data from ten healthy participants using motion capture and an instrumented treadmill, with protocols spanning continuous variations in locomotion tasks. It processed and screened the recordings into task- and stride-organized data.

Participants and sample planning

  • Ten healthy participants aged 20–60 years were recruited, with enrollment planning aimed for mean kinematic trajectories within five degrees of the target means.Eligibility required self-reported ability to walk over uneven ground and no lower-extremity joint problems or impairing neuromuscular disorders.Source: Participants · p. 2, Experimental Procedure · p. 3

Instrumentation

Walking

  • Walking trials varied speed across 0.8, 1.0, and 1.2 m/s and used multiple inclines; acceleration and deceleration trials tested 0.2 and 0.5 m/s^2.Walking speeds were normalized by leg length, and task order was randomized; inclines were 0°, 5°, and 10° in magnitude.Source: Walking · pp. 3–4

Running

  • Running trials covered 1.8, 2.0, 2.2, and 2.4 m/s, while walk-run transitions accelerated from rest to 2.2 m/s at 0.2 or 0.5 m/s^2.Two participants opted out of some running trials.Source: Running · p. 4

Sit-to-stand and stairs

  • Sit-to-stand trials measured force transfer between the stool and feet, while stair trials captured walk-stair transitions at four inclines and included no stair force plates.At least five ascent and descent trials were conducted at each stair incline; participants did not use their hands during sit-to-stand transitions.Source: Stair Details · p. 3, Sit-to-Stand · p. 4, Stairs · p. 4

Data processing

  • The pipeline segmented trials into tasks and strides, normalized strides to 150 points, and removed anomalous or fragmented strides; 8.3% of treadmill and 7.2% of stair strides were rejected.A median of 2006 strides per participant remained across modes and tasks, and no sit-to-stand trials were rejected.Source: Data Processing · pp. 4–5

Data Records

The dataset is distributed as two MATLAB structures containing continuous trial data and stride-normalized data.

Data Records

  • Streaming.mat contains continuous data from each trial, while Normalized.mat contains the same data parsed and normalized by stride.Further dataset documentation is provided in the corresponding README file.Source: Data Records · p. 5

Technical Validation

Marker placement was calibrated through treadmill walking, with checks for symmetry and expected joint range of motion.

Technical Validation

  • Marker placement was iteratively adjusted by less than one centimeter until two researchers judged joint symmetry and range of motion within expected ranges.The authors note that this calibration procedure would need modification for patient populations with different expected range of motion or gait asymmetry.Source: Technical Validation · p. 5

Usage Notes

The usage note points readers to an example MATLAB script for accessing the dataset and plotting a walking trial.

  • exampleUseScript.m demonstrates quick access to different data types in the MATLAB structure.Source: Usage Notes · p. 5
  • Its example plots sagittal-plane thigh kinematics and force-plate vertical load for AB04 walking at 1.0 m/s on a −10° slope.Source: Usage Notes · p. 5
  • The example superimposes all 59 strides from AB04’s one-minute walking trial.Source: Usage Notes · p. 5

Code Availability

The dataset and associated processing and treadmill-control code are available through MATLAB, with README files documenting their organization.

Figures & Tables

The figures depict locomotion modes, marker placement, and inter-participant kinematics across walking, running, sit-to-stand, and stair tasks. The tables identify sample-size calculations and parsed versus unparsed data formats.

Figures & Tables

  • Figure 1 distinguishes flat-surface walking and running from stair locomotion and depicts sampled speed and inclination variation, with Sit as the exception.Source: Figure 1 · p. 8
  • Figure 2 shows symmetric left- and right-limb marker placement, identifying markers used in the conventional gait model with an asterisk.Source: Figure 2 · p. 8
  • Table 1 is titled “Sample Size Calculations”; the accompanying method describes choosing a practical sample size within a range based on prior-experiment variability.Source: Table 1 · p. 8
  • Figure 3 presents inter-participant average walking kinematics and kinetics across inclines at 1.0 m/s, with shaded variation of one standard deviation.Source: Figure 3 · p. 9
  • Figure 4 presents inter-participant average running kinematics and kinetics at all tested speeds, with shaded variation of one standard deviation.Source: Figure 4 · p. 10
  • Figure 5 shows average sit-to-stand kinematics, while Figures 6 and 7 compare steady-state stair strides with ascent and descent transitions.Source: Figure 5 · p. 11, Figure 6 · p. 12, Figure 7 · p. 13
  • Streaming.mat contains data without parsing or normalization, whereas Normalized.mat contains data parsed and normalized by strides or cycles.Source: Table 2 · p. 14, Table 3 · p. 15
Loading 2108.12307v1…