Intelligent Assistive Technology and Systems Lab - click to go to homepage
IATSL develops assistive technology that is adaptive, flexible, and intelligent, enabling users to participate fully in their daily lives. Learn more about our research

Visit us:

Room 438

500 University Ave.

Toronto, Canada

P 416.946.8573

F 416.946.8570

 

Send us mail:

160 - 500 University Ave.

Toronto, ON, M5G 1V7

Canada

 

email us!

 

Follow IATSL on Twitter

Projects

Automated Analysis of Functional Balance

Keywords: Balance assessment, computer vision, functional assessment


Overview of Research

Balance and gait in the older adult population need special attention since they affect their functional mobility and safety. Previous studies have shown that most older adults report some difficulties with balance and these difficulties associate with falls in this age group (Hausdorff, Rios, & Edelberg, 2001).

Consequently, there is a need for measures of gait to monitor the recovery status of older adults over time and to assess their balance. On the other hand, typical gait analysis used in the clinic or lab cannot reflect long term changes in balance since balance failures often take place in homes, where standardized assessment cannot be applied. Moreover, tasks performed in the gait laboratory or clinics do not always simulate normal daily activities of older adult as people may feel more distracted at home than in the laboratory. One strategy working to address this limitation is the development of an affordable automated tool to better understand rehabilitation in home, capture balance failures taking place in home and reduce the frequency of clinical visits. To this end, the proposed study aims to develop an automated balance assessment tool which performs longitudinal, quantitative analyses of functional balance in a person’s own home through the analysis of common activities and movements.

Objectives

According to the above introduction and the overall goal, specific objectives of this study are to:

  1. Determine a “Gold Standard” currently used by therapists to assess the balance of older adults who have suffered a stroke, and how these map onto common activities and movements that may be performed in the home.
  2. Develop an automated tool which will perform longitudinal and quantitative analyses of functional balance of older adults who have suffered a stroke through capturing and analysing functional activities known to reveal balance change.
  3. Validate the accuracy and investigate the generalizability of the automated tool in evaluation of balance on an independent and unseen population.
  4. Explore how this automated tool can be deployed in client’s home and how it captures expert determined levels of balance performance over long term.

Methods

Our system includes the Microsoft Kinect (Figure 1), which will be deployed in the home to capture functional activities known to reveal balance change (e.g. gait, sit-to-stands). Several movement features and kinematic quantities can be computed from recorded joint positions using the Kinect via processing of 3D skeletal sequence. Using the machine learning algorithms, we will design a classifier to translate the reconstructed kinematics onto assessments of motor quality that are consistent with “gold standard” functional balance measures.

Kinect being used for balance analyais

Figure 1. Using a Kinect camera to survey a home environment.

Several movement features and kinematic quantities can be computed from recorded joint positions using the Kinect via processing of 3D skeletal sequence. Using the machine learning algorithms, we will design a classifier to translate the reconstructed kinematics onto assessments of motor quality that are consistent with “gold standard” functional balance measures.

Figure 2 is an example of the stride length measured on a balance impaired individual after Total Hip Replacement surgery across the spectrum of recovery, from pre-operative to 9 weeks post-operative assessment. As shown in the figure, the measured gait parameters were all worsened in comparison with pre-operative values after the surgery and all started to improve from 1 week to 6 weeks following the surgery. The preliminary results demonstrated the feasibility of using this type of system (specifically a Microsoft Kinect sensor and custom developed machine learning algorithms) for inferring the kinematics of functional activity recorded over time and being able to differentiate the different stages of recovery post surgery.

Plot of gait paramaters used in balance assessment

Figure 2. Example of change in gait parameters captured using the balance assessment system [click to enlarge image].


Funding Sources

NSERC CARE scolarship


Research Team

Elham Dolatabadi, University of Toronto

Alex Mihailidis, University of Toronto

Karl Zabjek, University of Toronto

Babak Taati, Toronto Rehabilitation Institute