Slide 2 Whereas the posterior parietal cortex plans for movements, the motor areas of the frontal lobe decide on execution of that plan. If the movement plan is to be executed, the frontal motor areas transform it into motor commands that reach the spinal cord, activate motor neurons, and move the limb. The important motor areas of the frontal lobe include the primary motor cortex (M1), the premotor cortex (PM), the supplementary motor area (SMA), and the cingulate motor area. We know little about the function of the cingulate. This lecture will focus on the functions of M1 and PM.
Coding object position with respect to the hand takes place in the premotor cortex Slide 3
Slide 4 Neurons in the posterior parietal cortex encode target and hand position in fixation centered coordinates. In the premotor cortex, these two vectors are compared and a new vector is computed: a vector that points from the hand to the target. Finally, in the primary motor cortex, this vector is transformed into the motor commands needed to move the arm.
Slide 5 To test the idea that in the premotor cortex, cells encode target position with respect to the hand (i.e., kinematics of the movement, not the forces), an experiment examined the neuronal activity in the ventral premotor region (PMv) in a task where the animal fixated a point and objects were brought close to the arm. A neuron’s tactile receptive field was mapped by touching various parts of the arm. The tactile receptive field for one PMv neuron is shown in B.
On each trial, the animal fixated one of three lights. A 10 cm white sphere served as a visual stimulus that was advanced along one of four trajectories toward the arm. The response of this cell to the visual stimulus is shown in C. The cell discharged strongly only when the stimulus was presented at a location that was on the side of the tactile response. As the fixation point was changed, the discharge did not vary significantly. Therefore, the visually evoked response remained fixed to the position of the visual stimulus with respect to the arm, not the fovea. When the arm was moved to the left side, the response was now strongest for trajectory III. That is, the visually evoked response appeared to move with the arm, and not the eyes. This is what would be expected if the position of an object were coded with respect to the position of a limb.
Neurons in the premotor cortex are sensitive to location of the target with respect to the hand and not forces
The results describe above suggest that the ventral premotor cortex might be a good place to look for cells that represent planning of reaching movements. The hypothesis is that if one is planning to reach for an object, cells in this region might code for where the target is located with respect to the hand, that is, the displacement vector. Importantly, if the location of the target remains invariant with respect to the location of the hand, the representation of the vector should also remain invariant. This would predict that if a monkey were to make reaching movements from different start positions of the hand, what matters is where the target is located with respect to the hand and not where the arm is located in the workspace, or where the arm is located with respect to point of fixation.
Slide 6 To test for this, discharge of cells in PMv were recorded in a task where a monkey was trained to move a cursor on a video monitor by moving its wrist. The monkey held a device in its hand that was connected to a computer that translated the stick’s motion to cursor motion. When a target was shown on the screen, the animal moved his wrist to bring the cursor to the target. The interesting point was that the movements were performed in three different initial wrist configurations. When the wrist was in the pronated position, the muscles that were activated to move the cursor to the target at 45o were quite different than the muscles that were activated to make the same cursor movement with the wrist in the supinated position. So if the cells somehow reflected the muscle commands that were needed to make the movement, then their activity should change considerably when the wrist configuration was changed. On the other hand, if the cells were coding the target direction with respect to the current position of the end-effector, where the end-effector now is not hand position but the cursor position, then their discharge might be invariant to changes in arm configuration. Indeed, among nearly all the task related PMv cells that were found, discharge was related to the direction of the target with respect to the cursor and not affected significantly by changes to the arm’s configuration.
Neurons in the motor cortex are sensitive to forces that are involved in making a reaching movement
Slide 7 Cells in PMv as a population appear to encode a movement in terms of a displacement vector with respect to the hand. Such cells are rare in the primary motor cortex (M1). In M1, most cells change their discharge as the configuration of the arm is changed, despite the fact that the cursor on the screen is moving the same way as before. Therefore, in M1 cells begin to transform the plan of the movement from a displacement vector with respect to the hand to patterns of activity that are necessary for activating the muscles and moving the limb.
Some cells in M1 have a discharge that correlates with forces produced by arm muscles
Slide 8 In this experiment, a constant torque was applied to elbow and shoulder joints of the monkey’s arm. The animal is trained to maintain constant arm position. Therefore, muscles produce activity to counter the imposed torque. Discharge for two neurons is shown as a function of torque imposed on the arm.
During a movement, activity in the premotor cortex precedes activity in M1
Slide 9 Consistent with the idea that the premotor cortex represents the movement plan with respect to the hand and the motor cortex represents the movement commands in terms of activities needed to guide the muscles, the activity in premotor cortex tends to precede the activity in M1.
Stimulation of the motor cortex results in twitch-like movements
Slide 10 High intensity stimulation of almost any part of the cerebral cortex produces a movement. However, the primary motor cortex produces movements with the lowest levels of stimulation. During brain surgery, the cortex may be stimulated and the resulting movements can be recorded. Stimulation results in discrete, flick-like twitches of a single muscle or small group of muscles on the contralateral side of the body. Movements are never skilled movements. Rather, they are flexion or extension of a single joint. In this slide, we see the notes made by a neurosurgeon regarding the effects that were observed. The dark line is the central sulcus, and the region anterior to it is the primary motor cortex. Note how in the medial aspect of the motor cortex, stimulation causes movements of the arm or the fingers, and that in the lateral aspects stimulation causes mouth and face movements.
Slide 11 There is a somatotopic organization of the body parts in the motor cortex. Body parts that are close to each other (for example, fingers are attached to the hand, which is attached to the arm), are represented by neuron in the motor cortex that are also close to each other. In this slide, we see the movements that are evoked by stimulating an anesthetized monkey. There is a general trend for somatotopy: trunk more medial, jaw more lateral. However, movement of a given body part (e.g., digits) is evoked from multiple foci.
Slide 12 This schematic is a summary of the somatotopy in the primary motor cortex. The amount of neural tissue dedicated to control of a particular body part is drawn to scale. Therefore, much more neural tissue is concerned with control of shoulder/hand/digits/thumb motion than control of the leg/feet/toes.
Damage to peripheral nervous system causes re-organization of the motor map
Slide 13 Connections between motor cortex and muscles are not fixed. In the adult rat, motor representation in the primary motor cortex can change with in a few hours after a nerve supplying motor axons to the muscles attached to the vibrissae (nose hair, or whiskers) is sectioned. This branch contains no sensory fibers. With in hours after cutting of the motor nerve, the motor map adjacent to the vibrissae region grows and takes up the region which used to evoke vibrissae movement.
Mechanism of reorganization of the motor map
Slide 14 There is a system of excitatory intracortical connections between motor cortical output neurons. This connection is not usually functionally expressed because of the intracortical fibers also stimulate local inhibitory neurons. Adjacent cortical regions expand when preexisting lateral excitatory connections are unmasked by decreased intracortical inhibition. We don't know how cutting a peripheral nerve or changes in sensory inflow might influence this intracortical inhibition.
Amputation reorganizes the motor map
Slide 15 An individual was involved in an accident and his right arm above the wrist had to be amputated. A year after the surgery, the motor cortex in each hemisphere was stimulated and muscle activity was recorded in the contralateral biceps. Note that the motor map for biceps on the left hemisphere is larger than the right hemisphere. This is because the biceps motor map in the left hemisphere has grown to take over the hand/finger regions that are no longer needed.
Phantom limb pain may be related to a lack of balance in the motor maps
Slide 16 After amputation, the individual is likely to experience chronic pain in the missing limb. The pain is more common in the initial years after amputation, but may remain for many years. Patients with PLP tend to have an imbalance in the size of the motor map between the hemispheres. In the top row of this figure, the brains of 3 individuals with hand amputations who do not suffer from phantom pain are shown. In these individuals, the motor map for biceps is similar in size between the left and the right hemispheres. In the bottom row of this figure, we have the brains of 3 individuals who suffer from phantom pain. In these individuals, the biceps region is large in the hemisphere contralateral to the amputated hand.
Slide 17 The brain relies on “internal models” to predict sensory consequences of motor commands. Occasionally, the models are incorrect. For example, when you get a new pair of glasses, the visual feedback that you receive when you watch your hand move may be slightly different that what you are accustomed to. This results in an error that re-calibrates an internal model. The idea is that in phantom limb pain, the error is so large that the nervous system continuously tries to adapt and never succeeds. That is, may be the pain associated with the missing limb is related to the fact that the patient’s motor system sends commands to the missing limb to move it, but it does not receive feedback indicating that it has moved.
Slide 18 To test this idea, one can place a mirror so that the patient views their normal limb and then ask the patient to make symmetric movements like clapping or conducting an orchestra while looking at the mirror. The patient may feel that the missing limb is finally moving, and in some cases this produces relief of the spasms and awkward posture that was causing the pain. In a well controlled study, McCabe et al. (2003; Rheumatology) found that training for a few minutes a day significantly reduced the perception of pain.
A stroke in the motor cortex results in reorganization of the motor map
Slide 19 After a stroke in the primary motor cortex, there is weakness and paralysis in the contralateral musculature. A gradual return of some abilities often occurs in the following weeks or months. In humans, there is rarely complete recovery of function in distal musculature. In this slide, we see the motor map in the motor cortex of a monkey. The lesion destroyed 21% of digit and 7% of wrist areas. After 3 months, the area for digits has been reduced further. This is probably because after a stroke, the affected limb is used less. It is possible that this reduced use itself affects representations in the brain, resulting in further loss of limb regions in the motor cortex. Therefore, it may be possible to reverse some of the effects of the damage through a movement therapy that "forces" the patient to use the affected limb.
Rehabilitation through forced use of the affected limb restores some of the lost motor map regions
Slide 20 Rehabilitation can prevent loss of motor cortical zones outside the region of infarct: the animal in this study received extensive training after the infarct and we see that there was a prevention of the loss of hand territory adjacent to the infarct region. Functional reorganization in the undamaged motor cortex was accompanied by behavioral recovery of skilled hand function. Therefore, the undamaged motor cortex can play an important role in motor recovery by reorganizing itself and compensating for the damaged areas.
Constrained motion rehabilitation can restore function in humans long after a stroke
Slide 21 Volunteers who were on average 5 years post stroke had their non-affected arm put in a splint for 90% of waking hours during a 2 week period. On 8 days during this two week period they came into the laboratory and were asked to perform movements with their affected hand for 6 hours each day. Results showed that the function of the affected hand was significantly improved because of the therapy, and the improvement lasted for months after the splint was removed.
Motor learning produces change in the motor map
Slide 22 Learning to make a skilled grasping movement results in changes in the motor cortex. Adult monkeys were trained to grasp small food pellets. The motor cortex was mapped using microstimulation both before and after training. Regions of cortex that evoked movements in a particular limb part were noted. After training in the hard task (small well training), there was an increase in the size of the motor map associated with the digits. However, if the task was easy (large well), there was no change.
Summary of functions of PPC, PM, and M1:
Slide 23 Posterior parietal cortex is important for spatial localization of our body and objects around us. Here, neurons align proprioception of arm with vision of hand and compute hand position in eye-centered coordinates. Neurons also compute object position in eye-centered coordinates.
In the premotor cortex, the location of an object that is the goal of the reaching movement is represented with respect to position of the hand. Here, neurons subtract target position with respect to hand position and code the desired movement in terms of displacement of the hand.
In the primary motor cortex, neurons transform the desired movement to muscle activity patterns and send this command to the spinal cord.