Plasticity In Si Maps

2.7.1 Evidence from Animals Shows that SI Maps Can Change

Related to difficulties in demonstrating detailed somatotopic maps in humans are possible effects due to the mechanisms responsible for cortical reorganization in adults after injuries (see reviews in References 53, 64, 120, 154, 155, and 193) or use and experience-dependent manipulations.39,53,64,122,177,194,195,227,238 Of greatest significance are demonstrations that maps may be dynamic since reorganization can be immediate following local anesthesia64 or temporally timed synchronization of peripheral inputs.39,53 The expression of changes at any level will result in altered cortical somatotopy, whether the mechanisms responsible for short-term changes in maps are from alterations in synaptic weights in intrinsic cortical, subcortical, or combined circuits. Changes in maps, receptive fields, response latencies, and degrees of correlated activity all stem from likely alterations in the balance between excitatory and inhibitory network elements possibly operating through Hebbian mechanisms.53,64,154,193 Thus, contrary to the precise boundaries for restricted body regions reported in studies with anesthetized animals, functional representations may be more general, rather than entirely dedicated to specific parts of the body. The brain imaging data discussed below is especially relevant in showing that cortical representations may expand or contract depending on behavioral circumstances.

2.7.2 Evidence from Humans Shows that SI Maps Can Change

Several brain imaging studies support the notion of a reorganized SI in adult subjects following peripheral or central injuries, sustained somatosensory stimulation through training, habituation to repetitive, simultaneous stimulation, or local anesthesia protocols. Much of this literature relies on observations from EEG or MEG recordings that are projected to be from the postcentral gyrus. However supportive these results are for findings of plasticity in animals, the data is sometimes contradictory and possibly open to alternative explanations. Brain imaging data by itself also cannot identify the locus responsible for observed changes. Instances of immediate, behaviorally mediated changes are likely due to unmasking existing subthreshold cortical connections through mechanisms, such as disinhibition.117 However, patients with long-term amputations may show reorganizations based on more extensive anatomic changes, such as the axonal sprouting that has been observed in animals.67,108,118 Changes in the connections between neighboring somatotopic representations in brainstem structures may be especially important (reviewed in Reference 165).

2.7.2.1 Effects Following Injuries

Amputees often report abnormal, unpleasant perceptual experiences of phantom limb sensations. These phantom sensations frequently arise from stimulation of skin areas whose cortical representations are immediately adjacent to the deafferented region. One hypothesis offered to explain phantom sensations is a remapping of SI where activity in cortical regions, with intact peripheral input, expands and takes over processing within the deafferented regions. These might best be described as soma-totopic reorganizations.165 For example, following amputations of the hand/arm, sequential stimulation of different facial regions elicits an orderly progression of sensations across the phantom (reviewed in Reference 193). Magnetic source imaging shows that these perceptual changes also correlate with extension of dipole source localizations to include the face and cortical region for the deafferented limb upon stimulating the face.61,62,66,240 Direct evidence of augmented activity from neighboring cortex is presented in a PET study with two patients that experienced a supernumerary limb upon stimulating the dorsal or ventral trunk.126 In this study, significant rCBF in SI expanded ventrally more than 20 mm into the deafferented hand/arm region from the normally evoked rCBF in the trunk region.126 Similar effects are seen in patients with amputations restricted to a single finger. In those cases involving amputations of digit two, the dipole source locations from stimulating neighboring digits one and three invade the deafferented region within less than 10 days following surgery.228,229 Referred phantom sensations also occur in spinal injury patients. In these cases, trigger zones that evoke phantoms may not necessarily be represented in adjacent sites in the SI map. A recent fMRI study describes displaced activation foci separated by more than 1.6 cm when such patients receive stimulation in trigger zones on the arm that also elicit the referred sensations to parts of the trunk represented below the segmental level of the spinal infarct.165 An exciting aspect of these results is direct confirmation of earlier ideas that perceptual experiences are tied to dedicated parts of the SI map even when activating disparate parts of the map from one stimulation site. This study also proposes a more general hypothesis that SI map representations after longer-term injuries include the possibility of somatotopic reorganizations in subcortical structures.165

2.7.2.2 Effects Following Training and Experience

Numerous studies provide examples of a reorganized SI following use and training experience. These findings indicate that plasticity can develop from an existing network without hypothesizing effects induced from injuries. One early MEG study describes the changes following surgical separation of webbed fingers in two patients.163 The SI hand area changed from a disorganized region before surgical separation of the fingers to a more normal map with distinct source localizations for each digit. A further example of dramatic dependence of the SI organization on training experience was first reported in MEG images from string players.60 In individuals who learned to play as children, the right SI area devoted to the left fingering digits, 2 to 5, appears larger than in control subjects. The left SI hand area in the musicians had a normal spatial extent. A similar instance of an enlarged representation occurs in experienced readers of Braille, who read several hours a day.177 Braille readers show a larger distribution of low threshold transcranial magnetic stimulation sites over the motor representation for the reading hand. An experimental confirmation of similar reorganized cortex from long-term experience shows multi-month long persistence of enlarged MR signal distribution in primary motor cortex for the fingers used in a practiced movement sequence.122 These results suggest that prolonged training leads to an alteration in the proportion of cortex dedicated to processing information from the restricted part of the body engaged in the task.

2.7.2.2.1 Effects Manifested to Task Specific Parameters Various factors appear responsible for such behaviorally induced changes in somatosensory maps. However, explanations for contrasting results in different studies are confusing because of the diversity of training, stimulation, and testing paradigms. Unpredicted from prior studies showing increases, one MEG study reports that the current dipole strengths decrease to stimulating skin surfaces following 3 to 4 weeks of practicing a same/different tactile discrimination task.216 Contrary to the view of yet greater cortex for a practiced usage, one explanation for this reduction after intensive experience might be the efficient use of fewer neurons to perform a well-learned task.216 An alternative, technical basis for these anomalous results is that plasticity effects are explicit to a task, and therefore not expressed when using different somatosensory stimuli during training and post-test, imaging trials, despite touching the same skin surfaces.140

In a study showing greater responses, subjects practice for three days and are tested with the same 21Hz vibration against the same digit surface when evaluating for changes in SEPs.140 In this study, the test stimulus is always applied to digit 3. In two different training sessions, stimulation is simultaneous on digits 2, 3, and 4 or just on digits 2 and 4. Post-training testing shows smaller responses to stimulating the non-trained digit and greater responses when training involved all three digits. The results suggest that simultaneous stimulation serves to expand the cortical territory devoted to the trained areas of skin. Hence, the decreases in responses to testing digit 3 are due to shrinkage of its territory when training involved digits 2 and 4.

The demonstrated effects on map organization in SI may be even more dynamic in especially depending on behavioral context during testing. For example, after 4 weeks of training to discriminate the orientation of tactile stimuli applied synchronously to digits 1 and 5, the Euclidian distance between the source locations for these two digits expands nearly five-fold during recordings obtained while subjects actually perform the discrimination task, but the distance between the very same digits shrinks during passive stimulation when no attention is needed.16 These results differ from findings in animals109,194 in showing changes in SI representations only during task-relevant trials. Furthermore, similar shifts occur immediately in humans, without extensive training, merely by altering the focus of attention (see above References 18, 19, and 174). These latter findings imply that training engages similar mechanisms responsible for directing attention to the conditions of a task.

2.7.2.2.2 Effects Manifested through Co-Active Stimulation The immediacy of these attention effects on map organization suggests that demonstrations of plasticity might depend on the flux of activity within existing networks. One possibility is the distribution of synchronized responses across adjacent cortex, because simultaneous stimulation is one factor present in all cited studies where training modified representations. Synchronized responses occur within cortical representations for attended stimulation (see above Reference 219). Thus, any procedures that lead to potential synchronized activity, without attention or behavioral training, might alter map representations, and even tactile discrimination perfor-mance.86 Examples of transformed maps, without a specific behavioral task, occur in animals53,85 and humans242 from sustained passive and co-active stimulation of receptive fields. In the animal study, a fused and enlarged representation follows coactive stimulation of adjacent receptive areas. In the human study, shrinkage of Euclidian distance between the source localizations for the median and ulnar nerves follows 40 minutes of co-active, tactile stimulation of the fingertips for digits 1, 2, 4, and 5. An explanation consistent with both results is the expansion of the cortical territory for the co-activated digits 1, 2, 4, and 5 into the representations for the unstimulated digit 3. Hence, in the study with humans, when subsequently stimulating the original nerves, whose representation includes all five digits, there is an apparent shrinkage in the modeling of the source localizations because the expanded activity now collapses into the previously unstimulated representation, which results in dipole distributions that lie closer together.

2.7.3 Changes in Active Foci

The simple explanation that instances of enlarged dipole distributions reflect increases in the cortical zones devoted to overused somatosensory inputs requires some scrutiny. In animals, direct recordings from area 3b of SI illustrate map changes. MEG and EEG studies present indirect evidence collected from remote electrodes, and through inverse modeling, map characteristics are projected back to the underlying cortex. In these experiments, could the map changes expressed in the dipole data reflect shifts in dominant activity to different cortical areas? This concern may be unnecessary as most studies report no changes in dipole orientations after experimental manipulations. A change in dipole orientation would be expected if an active focus occupies a new location. However, we describe a shift in the site of maximal blood flow changes from anterior SI (3b?) to posterior SI upon repeated trials (just four runs) of the same stimulus.28 The distance between peaks was under 10 mm and focal increases in blood flow still occur across multiple parts of SI. The change is in the location of maximum blood flow changes, which could appear as an enlarged SI representation in a dipole map, but without a change in dipole orientation. Other examples exist of even greater shifts in active foci following training (e.g., Reference 192). Many MEG and EEG studies demonstrating short-term manipulations of SI organization ignore the effects implied by these PET data and make no distinctions regarding plasticity effects in different parts of SI or other somatosensory regions. Direct evidence of expanded cortical activity is needed like that reported in patients with limb amputations126 or spinal injury.165

0 0

Post a comment