Robert E. Gander

     Robert E. Gander  B.Sc., M.A.Sc., Ph.D.

    Professor Emeritus of Electrical and Computer Engineering (1984-2011)

    Biography

    B.Sc.(Alta.)
    M.A.Sc.
    Ph.D.(Tor.)

    Dr. Gander's research interests are in the areas of instrumentation and biomedical engineering. In biomedical engineering, his primary interests are in signal processing, instrumentation, modelling of locomotion, and applications of artificial neural networks.

    His work in biosignal processing has involved myoelectric signals and nerve conduction velocity. In particular, power spectral analysis was used to determine the underlying characteristics of the myoelectric signal. This led to the development of a median frequency estimator that could be used for muscle fatigue monitoring. More recently an artificial neural network was used to classify muscle contraction levels after the myoelectric signal was converted to a phase-space domain. He has also worked in the area of determining the distribution of nerve conduction velocities in peripheral nerves using induced compound action potentials.

    Recent work in instrumentation has included a series of projects involving the measurement of sport activities. Starting blocks used by sprint athletes have been instrumented to measure horizontal and vertical force components for each foot. In addition, a Doppler microwave technique ("radar gun") is used to measure the athlete's speed for the first 20 meters after the start. Force-time and speed-time profiles are then provided as immediate feedback for the coach and the athlete to improve technique during a training session. A similar prototype was developed for long jump. Current work involves instrumenting standard rowing ergometers used in dry land training by rowers. Other current instrumentation projects involve adapting 2D ultrasound probes for 3D image acquisition, immobilized enzyme electrodes for in vivo monitoring of the immediate effects of stroke, and profilometry with potential application to ultrasonic imaging of breast cancer.

    A model of locomotor control has been developed that makes use of model predictive control applied to an active impedance model of the muscles used in locomotion. A single joint system has been successfully modelled based on the human knee. The model includes input for external disturbances that occur through actions at other joints particularly due to ground reaction forces. Both unidirectional tracking motions and cyclic trajectories based on human walking have been implemented. Earlier work developed a hierarchy of neural networks capable of generating trajectories in joint space. These trajectories could be used as input trajectories to the model predictive controller.

    Teaching

    2008-2009

    • BIOE800.3 - Advanced Medical Instrumentation
    • CME495.3 - Capstone Design Project
    • EE495.3 - Senior Design Project
    • GE121.3 - Engineering Design

    2007-2008

    • GE121.3 - Engineering Design

    Professional Activities


    • professional engineer (Association of Professional Engineers of Saskatchewan)
    • member, Canadian Medical and Biological Engineering Society
    • senior member, IEEE