Measuring human arm’s mechanical impedance for assessment of motor function

• Main Author: Mousavi Hondori, Hossein.
• Other Authors: Ling Shih Fu
• Format: Theses and Dissertations
• Language: English
• Published: 2013
• Subjects: DRNTU::Engineering::Mechanical engineering::Bio-mechatronics; DRNTU::Engineering::Mechanical engineering::Assistive technology
• Online Access: http://hdl.handle.net/10356/51865
• Institution: Nanyang Technological University

Assessment of motor function has remained a challenge for years. This is because of the complexity of human brain and the subjective nature of the assessment. On the other hand the assessment has lots of applications such as assessment of functional capabilities in post-stroke rehabilitation for which several scoring methods such as Fugl-Meyer or NIH stroke scale already exist but in general they are subjective or qualitative while for clinic purpose these methods are widely used. Hence, this thesis aims to develop a new tool for assessment of human’s motor function with a focus on the upper limb. For that purpose, Simultaneous Sensing cum Actuating technology (SSA) will be used for measuring force, velocity, and mechanical impedance of the limb. Impedance measured at the hand point while the upper limb is performing a motion task is an indicator of quality of the limb’s function. Here an electrical motor is used as sensor-actuator which carries a mechanical load (human’s limb); because the motor plays the role of a sensor as well, if we calibrate the “Transduction Matrix” of the motor, we can measure the mechanical impedance of the load through measuring the electrical impedance of the operating motor. A number of apparatuses are designed and fabricated (one-DoF rotary and one-DoF linear). Experimental and theoretical procedures are conducted to obtain the transduction matrix for the DC motor and linear motor mechanisms. In the experimental process, the systems operate with a set of pre-known loads and the angular velocity is continuously measured. Concurrently, voltage and current signals are recorded. Correlation of electrical input and the mechanical output gives the transduction matrix. The theoretical procedure, involves popular torque-to-current and angular velocity-to-voltage models as well as a transient mathematical model for the DC motor that altogether they provide the transduction matrix model. Then the arm’s mechanical impedance is measured for a number of subjects (8 distinct individuals) in several conditions. It is shown that while performing the one-dimensional rotary task, the subjects adjust their arm’s mechanical impedance to synchronize their arm’s movement with the rotation of the wheel. Their impedance is measured in 30 consequent trials to investigate their adaptation; the subjects minimized their impedance as they learned the task. More over it is shown that the value of mechanical impedance becomes more consistent with practice. Use of the linear motor and the transduction matrix provided measurement of impedance during reaching motion. Four distinct subjects are tested and for all subjects the impedance decreases and it converges to probably a minimum value required for doing the task. Since it is practically not possible for human subjects to have zero impedance at their arm, they decreases the impedance to reach to a lower level than the beginning of the trials which confirms that the mechanical impedance of the arm will be optimized while adapting the reaching task. In order to measure the mechanical impedance in form of a complex number, sinusoidal perturbation is applied to the arm using the DC motor. Using Hilbert transform and correlating the vibration of hand and the force applied by the motor real/imaginary parts of the mechanical impedance of human arm are separated. The experiment is done for two different postures (proximal and distal) and the results are compared. The results show that for human’s hand/arm system damping increases with frequency. The imaginary part of mechanical impedance that contains mass and stiffness increases with frequency; at 5 Hz the impedance was negative and it increased at 6, 7, 8 Hz until approached zero near 10 Hz. This suggests that the system’s resonance frequency is around 10 Hz. Further experiments presenting EMG vs. mechanical impedance shows a correlation between the impedance and the EMG. In both impedance graphs and EMG graphs it was observed that the magnitude of the signals increased when the arm was stiffened; this comparison validates the impedance measurement method. Finally, to verify the effect of postures on the impedance of human arm, a primitive model is employed. The model compares the results of a series of simulated end-point stiffness to the experimental studies done by other researchers. The compassion shows in most postures the model can fairly simulate the reality.