Haptics for Gait Retraining

We are using haptic feedback, through vibration and skin stretch, along with visual and auditory feedback, to assist in training new walking strategies. The work is aimed at healthy subjects and osteoarthritis patients, in order to reduce joint loads and to prevent or treat osteoarthritis.

A. Literature Reviews

Biofeedback Literature Review Spreadsheet -- This literature review explores various studies which have used multimodal biofeedback. In particular, it looks at the number of feedback modalities and abilities of subjects to respond to large number of sensors.

Knee Adduction Moment reduction methods Literature Review Spreadsheet -- This literature review looks at different methods of reducing the knee adduction moment to aid in preventing onset and/or progression of medial compartment knee osteoarthritis.

B. System Implementation Block Diagrams

Block diagrams provide generic description of the multimodal feedback system that is to be implemented. Two additional diagrams, which look at how learning might be incorporated into system and how the specific system for OA patients, are included.

1. Generic System

2. Generic System with Learning

3. Proposed System for OA Study

C. Knee Joint Loading and the Knee Adduction Moment

Osteoarthritis is a widespread disease that commonly affects the medial compartment (inside) of the knee joint. It has been shown that shifting the loading from the medial compartment to the lateral compartment (outside) of the knee joint is effective for preventing and treating knee osteoarthritis.

The common way of measuring knee loading is the knee adduction moment. This value is calculated by taking the cross product of the ground reaction force (GRF) and the position vector from the knee joint center (KJC) to the foot's center of pressure (COP) and keeping the component of the computed moment in the direction pointing directly away from the surface of the patella (kneecap).

We are using multiple haptic devices to retrain patients and healthy subjects to walk with a reduced knee adduction moment gait in order to treat or prevent knee joint osteoarthritis.

D. Implementing Multiple Haptic Feedback Devices

We use vibration (VibrationImplementation) and skin stretch (SkinStretch) haptic feedback to influence different gait parameters.

The desired foot progression angle is indicated with two vibration motors - one on the medial side (toward the body) of the foot and one on the lateral (away from the body) of the foot. If the foot progression angle is too large the lateral vibrotactor is activated and if the foot progression angle is too small the medial vibrotactor is activated.

The desired tibia bone angle, measured from a vertical line, is indicated by one vibration motor placed on the lateral side of the knee. If tibia angle is too small the motor is given a constant vibration and if the angle is too large the motor is given three short vibration pulses.

The desired backbone angle is indicated by either a skin stretch device placed on the lower back or two vibration motors place on either side above the hips.

We use multimodal haptic feedback to train 4DOFs gaits. The 4DOFs are foot progression angle and knee angle (implemented with vibration motors), trunk angle (implemented with skin stretch device)

E. Experiment Setup

Haptic, visual, and auditory feedback is used in the experiments. We use haptic/visual/haptic+visual feedback for training toe, knee, and trunk sway angle. Above stick figure is given to subjects as visual feedback. This feedback can reinforce the haptic feedback for each location. Right now, it is unclear to what extent giving subjects this additional feedback will assist or hinder in learning new gait patterns. In addition, auditory feedback is used for controlling stride length.

Motion capture system is useful to calculate knee adduction moment and investigate movement.

F. Experimental Validation

We tested 10 subjects with the following protocol:

(1) Single parameter gaits

Subjects were train to walk with five new gaits in separate trials. These gaits included (a) toe out, (b) toe in, (c) knee in, (d) increased trunk sway, and (e) increased stride length. Each of these trials lasted for 50 steps and subjects were given haptic, visual, haptic & visual, or auditory (for stride length) feedback to learn these single parameter gaits.

(2) Multiparameter gaits

Subjects were trained to walk with three subject-specific multiparameter gaits in separate trials. These multiparameter gaits were generated as described in the next section (Predicting New Gaits). Subjects were given haptic feedback to control their foot, knee, and trunk angles and auditory feedback to control their stride length.

G. Predicting New Gaits

Method for Predicting New Gaits (Simplified to 1D)

We would like to determine the gait which achieves at least 30% KAM reduction with minimal changes to subject's baseline gait.

Two steps to our Newton's method-like approach:

(1) Initialization: To initialize the algorithm, we choose a point in the direction of correlation for each of the parameters. For example, if knee and KAM are positively correlated, the initialized value for knee angle will be negative for a reduction in KAM.

(2) Iteration: Based on the final gait of the previous trial, we create a linear model about that point and predict new gait in direction of minimum change.

Using the final gait and knee adduction moment reduction from previous trial, we create a linear model around this point. (A is an (m x 4) matrix consisting of m point)

New gait is in x-direction and is given by:

w is a scalar which weights the x vector to predict a gait with 30% reduction in KAM

G. Predicting New Gaits

(1) Reduction in Knee Adduction Moment with Multiparameter Gaits

Multiparameter gaits reduce KAM more than 30% on average.

In particular, multiparameter gaits reduce KAM more than single parameter gaits and more than gait selected by subject (p < 0.05) (Jason and Pete's previous KAM study)

(2) Multiparameter Gaits are Subject Specific

Gaits are subject specific.

For example, subject 12 required much more trunk sway to get similar reductions as subject 2.

Sample Multiparameter Gait Achieved by Subjects 2 and 12:

	Subject 2	Subject 12
Toe Angle	-14.2	-10.9
Knee Angle	-5.3	-6.4
Trunk Angle	-2.6	5.1
Stride Length (% Change)	-1.8	-12.0
KAM Reduction (% Change)	-33.7	-33.8

(3) Localized Linear Model Prediction of Knee Adduction Moment Reduction

We can loosely track knee adduction moment using our localized linear model.