ASU LIMB Lab: www.limblab.org

Quantitative assessment of locomotor and upper-extremity function following neuromotor impairment

We seek to use our understanding of the biomechanics and motor control of movement to develop new approaches for quantitative assessment of motor function during locomotion and skilled movement. These new approaches are important for evaluating techniques and therapies for treatment of neuromotor injury. Although many methods exist for functional evaluation of locomotion following SCI, these outcome measures often rely on subjective scoring or complex tasks that are difficult to interpret. Objective, kinematic measurements of joint movements during locomotion are possible using video analysis. However, given the wide range of deficits and the complexity of locomotion, functional evaluation can be difficult even in the controlled case of treadmill stepping.

Kinematic Analyses of Locomotion

We are developing and validating objective, quantitative measures of locomotor function based on kinematic measurements. We hypothesize that some important aspects of locomotion to capture are:

        1. Movement frequency
        2. Movement shape (at a single-joint level)
        3. Movement coordination (among joints)
        4. Movement consistency
        5. Higher-order coordination (among legs)

        To assess movement frequency, we use a technique based on the Fast Fourier Transform (FFT). For each joint internal angle, the power spectrum was calculated from 8 seconds of continuous stepping using the FFT, and the frequency at peak power (FPP) was identified. The mean and standard deviation of the FPP was calculated across all five sham animals as an indicator of the range of frequencies associated with the peak power during sham bipedal locomotion. We constructed a bandpass filter centered at the mean FPP with a width of two standard deviations. The FFT power spectra for all trials of each treatment group and each joint angle were passed through the filter, integrated, and averaged across the ankle, knee, and hip joints of each hindlimb. The resultant average value of the integrated FFT (IFFT) was used as an index of the degree that a hindlimb exhibited movement kinematics with frequencies appropriate for restrained bipedal treadmill locomotion in sham rats.
        To perform the joint-level comparison of stepping kinematics, we use an analysis based on the continuous wavelet transform (CWT; Burke-Hubbard, 1998). Mean angle trajectories for each joint are averaged across uninjured animals during bipedal stepping to yield characteristic movement trajectories for each joint. The angle trajectories are individually normalized to zero mean, and then the set of leg angle trajectories is shifted in phase to minimize the total angles at the beginning and end of the resulting set of trajectories (while maintaining phase relationships among the trajectories). Each angle trajectory is then fitted with a 10th-order polynomial. The resulting polynomials for each joint angle trajectory are used as mother wavelets. For each stepping trial, CWTs are calculated for each joint angle at a scale corresponding to the mean stride duration observed during stepping by uninjured animals. Positive CWT coefficients indicate a positive correlation between the angular movements of the joint and the wavelet - an index of the similarity of the stepping kinematics around that time point to those of uninjured animals. The mean-squared value of the positive wavelet coefficients (negative coefficients were set to zero) are calculated over the entire time series for each joint. These mean-squared values are summed across joints of each leg of a given animal to yield a "wavelet resemblance" (WR), considered a measure of the degree to which individual joints exhibited step-like kinematics over the entire trial. Step-like movements of larger amplitude lead to higher WR values than smaller movements. The WR thus does not reflect the coordination among leg joints.

To assess the coordination among leg joints during locomotion, we use an analysis based on the cross-correlation function. Cross-correlations among all pairs of joints are calculated over a time lag range corresponding to +/- the average stride period for uninjured animals. The peak correlations for each unique joint-pair are averaged to yield an index of joint coordination for the leg, the Peak Cross-Correlation (PXC). The PXC reflects the consistency of coordination among joints, even if joint movements are offset in time. The PXC is less sensitive to the extent to which the specific kinematic waveforms are similar to uninjured stepping apart from differences in range of movement. Consequently, although the IFFT, PXC and WR both increase for larger movements, they measure largely independent aspects of stepping kinematics.

Neurophysiological Assessment
        We are working to improve on noninvasive methods for analyses of spinal cord recovery after injury. Noninvasive electrophysiological methods have been useful for quantitative assessments of spinal connectivity following injury and evaluation of treatment strategies. Motor-evoked potentials (MEPs) can be elicited by stimulation of the motor cortex or spinal cord, even with acutely-implanted, subcutaneous electrodes. In humans, MEPs have been elicited by magnetic stimulation of the brain, or by electrically stimulating the brainstem using surface electrodes placed over the mastoid processes of the skull (Taylor and Gandevia, 2004). Similar techniques are also effective in animal models such as rats. For motor cortex stimulation in rats, the cathode has typically been implanted beneath the skin on the skull over the motor cortex, and the anode implanted in the nose (Lopez-Vales, et al., 2006). Brief (100 microsecond) relatively powerful (25 milliampere) electric pulses stimulate the brain, and the resulting evoked potential is recorded electromyographically from muscles in the limbs. Although gross measures, such as latency and amplitude of the resulting potentials have been reported for diagnostic purposes, relatively few studies have investigated the specific pathways contributing to MEPs, and the degree to which detailed analysis of the evoked potentials could yield information about spinal connectivity (Gennuso, et al., 1991). Consequently, we have initiated a study into electricaly-induced MEPs through cortical and spinal stimulation to gain a better understanding of how this technique can reveal neural plasticity following SCI.

Using reconfigurable technology for quantitative assessment of motor control in humans (in collaboration with Roozbeh Jafari)

We are developing a clinical movement assessment system (CMAS), a robust and flexible hardware and software system for quantitative measurements of motor effort following neuromotor injury. The system will allow for easy and versatile data collection by non-technical staff and patients from a variety of devices by using recent advances in technology for real-time reconfiguration. We are currently working to demonstrate the effectiveness of the system by developing two patient-friendly devices. One device measures individual finger forces in flexion and extension, and one device will measure finger movements for extended periods in a home setting. The effectiveness of the CMAS and finger devices will be demonstrated by using the system to make quantitative assessments of neuromotor effort following stroke in a patient population.
    This project will result in new techniques that use advances in system reconfiguration to create a versatile networked platform capable of supporting many devices to measure neuromotor function. The important considerations of security, power management, and fault tolerance will be fundamental elements of the system design. The system will include hardware and software features designed to accommodate a wide range of patient specific motor capabilities, and provide real-time output of dynamic forces and positions with visual and auditory feedback in an interactive interface.
     For neuroscience, the CMAS system will allow for unique measurements of finger force generation and coordination capabilities in flexion and extension following neuromotor injury. The CMAS system will also allow some of the first quantitative measurements of hand use for extended periods in the home environment.
The development of this system will have broad impact in basic neuroscience, computer science, clinical medicine, and rehabilitation. Standardized, quantitative functional measurements will contribute to understanding neuromotor injuries such as stroke or spinal cord injury. Quantitative measurements of motor function are the first step in developing appropriate population databases to enable rapid, patient-specific assessments of motor function. In addition to the two devices currently under development, emerging sensor technologies promise to enable the development of additional devices capable of measuring a wide range of neuromotor function. The CMAS will ultimately enable scientists and clinicians to assess the strength and accuracy of motor control of flexion and extension of the finger, hand, wrist, and elbow, pronation and supination of the forearm, and shoulder movements. This system can also be used for rehabilitative training. Moreover, this project will help computer scientists use this functional information to implement fault tolerance, power optimization and coordination strategies among various devices and sensors. Use of these devices in clinical practice will contribute to improved prospects for diagnosis and rehabilitation of injured individuals.