Hindlimb Motor Responses Evoked By intraspinal microstimulation In The anaesthetized and acutely prepared cat

Dorn Spinal Therapy

Spine Healing Therapy

Get Instant Access

4.3.1 Background

The long-term goal of our work is to assess the feasibility of motor system neural prostheses based on electrical activation of spinal neural networks. In this section we report on two series of experiments designed to characterize the hindlimb motor responses evoked by intraspinal microstimulation of the lumbar spinal cord in the cat. In the first series of experiments, high-resolution mapping was used to identify the areas of the spinal cord that produce motor responses when stimulated. The second series of experiments was designed to characterize further the hindlimb motor response evoked at the identified locations. For this purpose, measurements were made of the end-point force evoked at the paw at a number of hindlimb positions in the workspace pattern. The results indicate that microstimulation at some regions in the lumbar spinal cord can generate organized, multiple-joint, multiple-muscle motor responses, which are not obtained by direct stimulation of muscles or by intraspinal stimulation of motor neurons.

4.3.2 High Resolution Spatial Maps of Knee Torque Generation

Experiments were conducted to map the hindlimb motor responses evoked by intraspinal microstimulation of the lumbar segments (L5-L7) in neurologically intact, chloralose-anaesthetized adult cats. The isometric torque generated about the knee joint and intramuscular EMGs from knee flexors and extensors were recorded in response to intraspinal stimuli (1s, 20Hz, 5-100 pA, 100 ps) applied with metal microelectrodes. Stimuli were applied in ipsilateral and contralateral segments with a 250 pm mediolateral resolution and a 200 pm dorsoventral resolution to create fine grained maps at half-segment intervals (i.e., L6, L6/L7 boundary, L7, etc.).

* By Warren M. Grill and Michel A. Lemay.

Anaesthetized Cat

FIGURE 4.3 Isometric knee extension and knee flexion torques evoked by intraspinal microstimulation of the ipsilateral L6 spinal cord in a chloralose-anaesthetized cat (50 |A, 100 |s, 20 Hz, 1 s). The ordinate corresponds to the midline of the spinal cord; the bold lines indicate the dorsal surface of the spinal cord and the ventral extent of the mapped region. Torques were mapped with a dorsoventral resolution of 200 |m and a mediolateral resolution of 250 |m, and are plotted as continuous grey-scale values through triangular interpolation.

FIGURE 4.3 Isometric knee extension and knee flexion torques evoked by intraspinal microstimulation of the ipsilateral L6 spinal cord in a chloralose-anaesthetized cat (50 |A, 100 |s, 20 Hz, 1 s). The ordinate corresponds to the midline of the spinal cord; the bold lines indicate the dorsal surface of the spinal cord and the ventral extent of the mapped region. Torques were mapped with a dorsoventral resolution of 200 |m and a mediolateral resolution of 250 |m, and are plotted as continuous grey-scale values through triangular interpolation.

Torque maps were repeatable across experiments with a strong congruence between the spatial locations where torques were evoked and similar relative magnitudes of flexion and extension torques. Maps of the torques evoked by intraspinal microstimulation revealed a rostrocaudal, mediolateral, and dorsoventral organization of regions of the spinal cord that evoked motor responses about the knee joint (Figure 4.3). At L6 and L7, stimulation in the ipsilateral dorsal aspect of the cord produced strong flexion torques that accommodated rapidly during the 1 s stimulus. Weaker flexion torques were also produced by microstimulation at the lateral aspect of the intermediate region/medial aspect of the white matter. Stimulation in the contralateral dorsal horn produced extension torques that accommodated rapidly. The spatial maps of torques evoked by microstimulation in the ipsilateral ventral horn varied at different rostrocaudal levels. The spatial representation of flexors in the ventral horn increased at more caudal levels as compared to more rostral levels, and there were differences in absolute torque magnitudes. For example, stronger flexion torques were produced by microstimulation over a larger area in L7 than in L6. In the lateral ventral horn, microstimulation produced strong extension torques, while microstimulation at more ventral and medial locations produced flexion torques. Thus, extension torques were produced at locations that were lateral to locations producing flexion torques. EMG recordings indicated that microstimulation in the ipsilateral dorsal, intermediate, and medial ventral locations activated knee flexors selectively, while microstimulation in lateral ventral locations activated knee extensors selectively. However, myostimulation at 50-100 pA did not provide selective activation of individual synergists. From L5 though L7 there was a large area in the intermediate region of the cord which produced no torques (Figure 4.3).

4.3.3 Characterization of Limb Motor Response

Fine mapping of the knee joint response evoked by intraspinal microstimulation identified regions of the spinal cord that produce hindlimb motor responses and established surgical and stimulation methods in the cat model, but these results provided only a limited picture of the limb motor response. As demonstrated by Giszter et al.10 in the frog, measurements of endpoint forces provide a more complete representation of the limb responses to stimulation, and with careful measurements of the limb geometry, still allow recovery of the individual joint torques.

The end-point forces elicited by intraspinal microstimulation (0.5-s train of 40-Hz 100-pA 100-ps biphasic current pulses) of the L5-L7 spinal cord were measured in adult cats either anaesthetized with chloralose or decerebrated. End-point force vectors in the sagittal plane were measured at 9-12 positions of the hindlimb that spanned the range of knee and ankle angles encountered during locomotion. Bifilar wire electrodes were inserted into four hindlimb muscles (knee flexor, knee extensor, ankle flexor and extensor) to record the EMG activity: biceps femoris, vastus medialis or lateralis, tibialis anterior, and medial gastrocnemius, with electrode location verified via post-mortem dissection. The raw EMG signal was amplified, filtered (10-1000 Hz), and sampled at 2500 Hz. The animal's pelvis and femur were held with bone pins, and the paw was attached to a gimbal mounted on a movable six-axis force transducer. The kinematic linkage was thus knee and ankle as opposed to hip and knee as in the previous experiments in the frog. Force-fields were constructed by dividing the workspace into triangles and estimating the force vectors within a triangle by a linear interpolation based on the vectors measured at the vertices (Figure 4.4). We limited our analysis to the sagittal plane since this is the primary plane of limb motion during locomotion. Forces were divided into a passive component (force measured before the onset of activation) and an active component (total forces measured minus the passive portion).

The number of response types elicited via intraspinal microstimulation was limited, similar to what has been found in the frog10 and rat.53 With ipsilateral stimulation, force fields (FFs) were of four types (Figure 4.5): flexion withdrawal, rostral flexion, caudal extension, and rostral extension. Flexion withdrawal was the most common field type encountered (33 of 54 fields), especially at superficial depths of stimulation (800 pm), and is probably the traditional flexion withdrawal reflex.1 Extensor fields were most commonly encountered at ventral depths (5000 pm), and appeared to be the result of direct activation of motoneurons. FFs found at intermediate depths (1500 pm) presented the largest variety.

The characteristics of the FF differed between ventral vs. dorsal and intermediate locations (Figure 4.6). Endpoint forces at dorsal or intermediate locations varied both in direction and magnitude with changes in the limb's configuration. For some stimulation sites, the end-point forces varied so as to produce a FF which converged

FIGURE 4.4 Endpoint forces evoked by intraspinal microstimulation were measured at 9-12 different positions of the hindlimb shown in the left panel. With the femur held fixed via pins, the knee and ankle were moved through a large range of motion (joint range of motion during walking shown by the thick arcs). Force-fields (FFs) were constructed from measured force vectors (dark arrows) by interpolation (light arrows). Forces represented are the total forces (active and passive) minus the component due to gravity. The FF evoked by stimulation at 1500 jam exhibited a point of convergence in the workspace, where the net endpoint forces were zero, while the FF evoked at 5000 jjm did not.

FIGURE 4.4 Endpoint forces evoked by intraspinal microstimulation were measured at 9-12 different positions of the hindlimb shown in the left panel. With the femur held fixed via pins, the knee and ankle were moved through a large range of motion (joint range of motion during walking shown by the thick arcs). Force-fields (FFs) were constructed from measured force vectors (dark arrows) by interpolation (light arrows). Forces represented are the total forces (active and passive) minus the component due to gravity. The FF evoked by stimulation at 1500 jam exhibited a point of convergence in the workspace, where the net endpoint forces were zero, while the FF evoked at 5000 jjm did not.

to a point in the workspace where the net active end-point force was zero (Figure 4.4). Electromyographic activity of four hindlimb muscles (tibialis ant., med. gastroc, bicep fem., vastus med. in the example in Figure 4.4) indicated that the convergent FFs were produced by coactivation of multiple muscles. In contrast, while the magnitude of end-point forces evoked by ventral microstimulation varied with limb position, their directions were largely invariant, and the resulting FFs were parallel or divergent (Figure 4.4). Similar parallel or divergent patterns were evoked by intramuscular stimulation of individual muscles.

Was this article helpful?

0 0

Post a comment