UNSW

Estimating Lower Limb Kinematics using Distance Measurements with a Reduced Wearable Inertial Sensor Count

Luke Wicent Sy, Nigel Lovell, Stephen Redmond

Gait Analysis

Osteoarthritis

Osteoarthritis

Cerebral Palsy Surgery

Cerebral Palsy Surgery

Parkinson's Disease

Parkinson's Disease

Perform. Improvement

Perform. Improvement

Fall risk assessment

Fall risk assessment

Real Time Feedback

Real Time Feedback

Motion Capture Systems

Camera based

Camera based

Very accurate but limited to a small space

IMU based OSPS

Inertial Measurement Unit (IMU)

Miniaturization. Track position and orientation (albeit with drift).

IMU based OSPS

IMU based (one sensor per seg)

Can capture almost everywhere. Can be conspicuous for everyday use

Wearable based sparse Wearable based sparse

Wearable based

More comfortable

  • Soft stretch sensors
  • More & smaller IMUs
  • Sparse sensors

Wearable based

More comfortable

  • Soft stretch sensors
  • More & smaller IMUs
  • Sparse sensors

Sparse Wearable Challenge

Goal: Comfortable, Fast, and Accurate Motion Capture System

Camera based Missing sensor

One sensor per segment.

Less sensor = Missing info

Infer through biomechanical constraints

Infer through biomechanical constraints

Infer through biomechanical constraints

Infer from additional measurements

Sparse Constrained KF (CKF)

Algorithm overview of prior work

Sparse CKF Sparse CKF

Body Model:

CKF+D Body Model

Pred.

Sparse CKF Prediction

Meas.

Sparse CKF Measurement

Cstr.

Sparse CKF Constraint

Sparse CKF - Sample

CKF Sample Walk

Sparse CKF - Weakness

To make it work, we need to make assumptions that may not be practical for certain movements (e.g., Activities of Daily Living or ADLs).

 

CKF Sample High Knee Jog CKF Sample Jog

Sparse CKF + Distance

Overview

Distance measurement which can be obtained through ultrasonic or ultra-wide band radio (UWB).

Sparse CKF

Body Model:

CKF+D Body Model

Sparse CKF + Distance

Remove pelvis related assumptions. Additional measurement model.

Sparse CKF

$$ \mathbf{H}(\mathbf{x}) = \tau_{k}^{pla} ( \hat{\theta}_{lk} ) \\
\tau_{k}^{pla} ( \theta_{lk} ) = \overbrace{\tfrac{d^{p}}{2} \mathbf{r}^p_y - d^{ls} \mathbf{r}^{ls}_{z}}^{\psi, \text{ hip + shanks}} + \overbrace{d^{lt} \mathbf{r}_x^{ls} \sin{(\theta_{lk})} -d^{lt} \mathbf{r}^{ls}_z \cos{(\theta_{lk})} }^{\Lambda, \text{ thigh}} \\
(\hat{d}^{pla})^2 = \tau_{k}^{pla}(\theta_{lk})^2 = \psi^2 + 2 \psi \cdot \Lambda + (d^{lt})^2 \\
\alpha \cos{(\theta_{lk})} + \beta \sin{(\theta_{lk})} = \gamma \\
\alpha = - 2 d^{lt} \psi \cdot \mathbf{r}^{ls}_z, \quad \beta = 2 d^{lt} \psi \cdot \mathbf{r}^{ls}_x \\
\gamma = (\hat{d}^{pla})^2 - \psi^2 - (d^{lt})^2 \\
\hat{\theta}_{lk} = \cos^{-1}\left( \tfrac{\alpha \gamma \pm \beta \sqrt{\alpha^2+\beta^2-\gamma^2}}{\alpha^2+\beta^2} \right) $$

Sparse CKF+D - Sample

Tested on actual IMU data + simulated distance measurement from Sparse CKF dataset (walking, jumping jacks, speedskater, TUG, jog). Most deviation is at the turning motion ($t=3.5 - 5$s).

 

Sample Walk Varying sigma

Sparse CKF+D - Sample

Dramatic increase in performance in dynamic movements. Captures Sagitall knee angles better.

 

Sample High Knee Jog Sample Jog

Sparse CKF+D - Sample

Is able to locate relative position better. Note: TUG = Timed Up and Go.

 

Sample SpeedSkater Sample TUG

Sparse CKF+D - Varying $\sigma$

Simulated at different levels of distance measurement noise $\sigma$ (assumed gaussian). Useful from $\sigma \leq 0.1$ m for walking. Useful from $\sigma \leq 0.2$ m for dynamic movements.

Varying sigma

Conclusion & Future Work

  • Adding distance measurement is indeed a promising approach
    • Can be implemented using ultrasound or Ultrawideband (UWB) based sensors.
    • Simulated distance measurement needs actual validation.
  • Code at https://git.io/JvLCF
  • Interesting to try with better models and tracking more segments.
Infer from distance measurements LG7Seg Sample Walk

Infer from distance.   Lie Group based 7 segment

Appendix - CKF+D Quant. Results

Tested on actual IMU data + simulated distance measurement from Sparse CKF dataset (walking, jumping jacks, speedskater, TUG, jog).

Varying sigma