Body, action, and space representations in people affected by spinal cord injuries

List of abbreviations

BR

body representation

PPS

peripersonal space

RT

response times

SCI

spinal cord injury

Introduction

Anecdotical reports concerning modifications of the perception of the body, space, and action following a spinal cord injury (SCI) are well known to professionals that have experienced with these patients as they can report body misperceptions (e.g., phantom limb or body loss), misrepresentations of space (e.g., body or limbs position, distance lengthening), changes in action representation (e.g., illusionary and phantom limb movement leading to muscular fatigue) ( ; ; ; ). These sensations, linked to pain, the level of the lesion and the time passed since the lesion onset ( ), are symptoms of neuroplastic changes in areas related to the body, space, and action representation.

A spinal cord interruption, especially when this is not developmental but acquired, abruptly changes the brain–body communication. This modification alters the balance between the body sensorial inputs, the signals sent toward it, and the brain activity, leading to abnormalities in cortical and subcortical activity. Consequently, the brain neuroplastically changes itself ( ) into brain networks that are involved in body, space, and action representations. Therefore, modifications in these cognitive functions are likely to occur.

This chapter is organized as: a first section reviewing the main literature concerning neuroplastic modifications in SCI individuals, underlying those related to the body, space, or action representations, the following three sections will show the experimental outcomes from studies concerning these representations in SCI individuals, and a fourth section will highlight the experimental results showing the effects of rehabilitative trainings on these representations.

Finally, a concluding section will present the main findings, interpretations, and connections.

The neuroplasticity following SCI and the networks involved in the body, space, and action representations

Neuroplasticity mechanisms have been recently investigated in SCI. Different levels of activation in the sensory and motor cortices have shown, particularly a reduction of activity in areas corresponding to paralyzed limbs ( ) along with increased activation in areas corresponding to spared body parts ( ; ).

Other studies ( ) showed the expansion of the cortical hand area toward the leg area during upper limbs movements. An enlargement of the face representation in the surrounding areas in monkeys after SCI was observed in the primary and nonprimary somatosensory cortices ( ).

Other works found a general reorganization of somatotopic areas ( ; ) and their degeneration over time ( ).

All these neuroplastic changes relate to the body, space, and action representations; indeed, all the somatotopic areas are involved in BR ( ). Moreover, in BR, the posterior parietal cortex, integrating multisensory inputs and outputs has a fundamental role ( ) as this brain area is directly involved in the neuroplasticity following SCI, because of the sensorial modifications.

The understanding of others’ actions and space representations rely also on somatosensory cortices. In fact, the fronto-parietal “mirror” network ( ) includes the ventral premotor cortex and the inferior frontal gyrus. Space perception involves the intraparietal sulcus, the lateral occipital complex, and the premotor cortex ( ; ) in the peripersonal space representation, while the extrapersonal space perception is linked to an occipito-parietal circuit, a parieto-prefrontal pathway, and a parieto-premotor one ( ).

The unbalance between the somatosensory and other sensory feedbacks, causing fundamental changes in the neurological system after SCI, might cause of maladaptive neuroplasticity in these cognitive representations, causing modifications that may be subtle, or evident like phantom sensations ( ) and neuropathic pain ( ).

The body representations

Body ownership is an important aspect of our body representations (BR) and can be defined as the sensation that a bodily part is part of our body. Relevant disturbances can lead to Asomatognosia ( ), namely abnormalities in various facets of body ownership such as the experienced existence, visual self-recognition, and sense of belonging to contralesional body parts.

This facet of BR, investigated by the rubber-hand paradigm, appears to be modified according to neuroplastic cortical modifications. In the commonest version of the rubber-hand illusion (see Fig. 1 , ), participants observe a rubber-hand, i n a congruent position with their own body, synchronously touched with their hand, hidden from sight ( ). An asynchronous rubber-hand/real-hand touch should not elicit a body ownership sensation and, therefore, is used as control condition ( ).

In hand representation-deprived tetraplegics, the ownership sensation toward a rubber hand can be elicited by the synchronous tactile stimulation of the left cheek and the rubber hand (see Fig. 2 ; ). This is probably caused by the overlapping of the cheek somatic and motor representations over the representation of the hand that are normally contiguously represented ( ). A similar result with synchronous visuo-tactile stimulation of the upper back and rubber legs in SCI individuals with complete paraplegia was found ( ). Moreover, the subjective verbal report of the sensation of body ownership toward a body part seems to be captured by the mere vision of the body part, especially when there is a total lack of tactile sensations (lower limbs in people affected by complete paraplegia: ; rubber hand in a patient affected by tetraplegia: ).

A further, implicit paradigm to study body ownership relies on the body-view enhancement effect. This task exploits the effect of visual stimuli appearing on our body that evoke faster reactions than visual stimuli appearing on the same space location, but on a neutral surface ( ). By a modified version of this paradigm, it was observed that individuals affected by complete paraplegia feel body ownership for the parts of the body that have spared sensory and motor functions, but also the same level of body ownership for their wheelchair, while the lower limbs show no body ownership ( ). Interestingly, the same task executed with the SCI participant seated in a different, never used before, wheelchair shows that this redefines their whole BR: there is no body ownership toward the new wheelchair, the lower limbs, nor for the part of the body that has spared motor and sensory functions ( ).

The notion of body schema dates back to the seminal work of and can be nowadays defined as the dynamic representation of body parts and their movements, which derives from multiple sensory and motor inputs (e.g., proprioceptive, vestibular, tactile, visual, efference copy). Patients with an impaired body schema are not able to distinguish the position of their arm when their eyes are closed.

A popular experimental paradigm investigating this aspect is the Mental Body Rotation paradigm ( ). According to this, individuals are asked to identify the laterality of a hand or a foot presented on screen at different rotations. The underlying mechanisms of decision imply that participants rotate the bodily stimuli congruent with their body and, therefore, response times (RTs), required to align series of gradually more rotated images to the vertical congruent with the body position, are slower. This leads to a typical RTs pattern, characteristic of a normal body schema representation, called bio-mechanical effect ( ).

Mental Body Rotation paradigm confirms that complete SCI disrupts the influence of postural changes on the representation of the deafferented body parts (feet, but not hands) and, regardless the posture, whole-BR progressively deteriorates proportionally to SCI completeness ( ), as a direct effect of the somato-sensation lack in the affected body-parts in SCI individuals.

Indeed, in SCI participants, the misproprioception of the paralyzed body parts is not uncommon (77% of cases) and it is present in all people affected by complete tetraplegia ( ).

Also Disownership-like and Somatoparaphrenia-like sensations (sensations of not owning their own limbs, even attributing them to someone else, as in the latter case) are often reported (48% and 46%, respectively), altogether to illusory motion of paralyzed body parts (51%).

In all these cases, these bodily illusions are connected to the time since the lesion onset, the level of injury, and the presence of pain ( ).

Interestingly, these bodily illusions are less evident in presence of neuropathic and visceral pain. This apparently counterintuitive result addresses toward the fact that, in a deafferented/deefferented body, a pain sensation coming from the impaired body part is used by our cognitive system as a surrogate BR.

The action representation

The representation of actions is fundamental in our everyday life. It allows us to plan our gestures, interact with the environment and other people. The representation of actions is based on the fronto-parietal “mirror” network ( ; ; ; ).

In matching-to-sample experiments, participants are presented with a single stimulus called the sample and then with two choice stimuli called the comparisons. The participants have to select the comparison that matches the sample. A version of this paradigm that used images of bodies that could be different for body action or body form, for the whole body, in its upper part, or in the lower part, was applied to a sample of complete paraplegics and healthy controls ( ). Results showed that complete paraplegics participants had slower reaction times and greater errors incidence, in body images that were different for the lower part of the body, than the control group of healthy participants.

These results were confirmed and widened by using a Temporal Occlusion Paradigm ( ) where participants observed two sets of videos depicting an actor who attempted to climb onto a platform using a wheelchair or rollerblades. Each video could end in three different ways: (a) successfully (the actor went up the step), (b) failing (the actor stopped before the step without going up), or (c) falling (the actor fell without going up). Each video was cut in five different durations at 600, 1200, 1800, 2400, and 3000 ms. In the shortest (600 ms) clip, only preparatory body movements were shown, while in the longest (3000 ms), the complete action (see Fig. 3 ).

Participants had to report the outcome of the video choosing among success, fail, or fall. Participants were divided in three groups: the complete paraplegics group, the expert skaters group, and a group composed by physiotherapists that had at least 1 year of experience in a spinal-unit, but no experience in rollerblades or other skating tools.

Results from this study showed that complete paraplegics participants had a worse performance in rollerblades videos compared to the Skaters group, but also compared to the Physiotherapists group, inexperienced in skating (see Fig. 4 ), suggesting that abnormalities in the mirror system for the lower limbs affect the recognition of lower limbs’ actions in a subtle, but still worsening way compared to a group that had no direct experience in the action.

A further indication regarding modifications in the action representation system derives from the tendon-vibration illusion, able to induce illusory perception of movement ( ). The application of this illusion to individuals affected by SCI induced movement illusions that were qualitatively different from the healthy control sample (e.g., as if the arm wanted to extend itself, or a sensation of pushing against something) that may reflect different reorganization processes following SCI ( ).

The modifications in the action representations could also affect the learning of new motor sequences. In fact, according to Bloch and colleagues ( ), it exists a specific deficit in implicit learning of motor sequences in SCI people, and it may be extended even in the motor learning abilities consequently a SCI lesion.

However, complete SCI participants demonstrated a better performance in forecasting the outcome of wheelchair action in a Temporal Occlusion Paradigm than Physiotherapists working in a Spinal Cord Unit and a group of expert skaters ( ). Interestingly, experimental results confirmed that agonistic wheelchair-basket players showed a better performance with images changing only in the upper-part of the body than control healthy agonistic basketball players ( ); moreover, a greater EEG observational contingent negative variation waveform, considered a marker of action effect prediction, during the evaluation of free throws videos, than control participants ( ). Finally, with a different experimental paradigm could not replicate the impairment of implicit motor learning previously reported. Therefore, the motor learning impairment in SCI is probably not present, or it is very specific for some motor nonecological tasks.

Another aspect of action representation is the motor imagery. This can be defined as the act to internally represent an action, without its execution ( ). Nonetheless, motor imagery involves the activation of brain structures involved in action performance, such as premotor areas and the left intraparietal sulcus, even if at a minor level than actual action performance ( ; ; ).

While in some studies, motor imagery seems to be spared after SCI ( ; ), altered cortical activations were found ( ; ).

In particular, during motor imagery tasks, the primary and nonprimary motor cortices, and subcortical structures are more activated in individuals affected by SCI than control subjects ( ), even if the imagery does not involve paralyzed limbs ( ). Similarly, the minor activation of the primary motor cortex shown by healthy individuals during imagined movements compared to real movements, it is not present after SCI ( ), suggesting that after SCI, the motor cortices are activated in the same way during motor imagery or real movements.

Experimental reports suggest that these effects lead to a different motor imagery strategy after SCI for subjective responses ( ). In fact, from behavioral investigations, motor imagery after SCI appears to be somato-topographically organized, with higher ability for the spared body parts, and it is impaired in presence of neuropathic pain ( ).

The space representation

The study of the space representation, since the seminal works by Rizzolatti and colleagues ( ) was divided in extrapersonal and peripersonal space, two spaces that can be broadly defined respectively as the space outside and within the reach of our limbs.

PPS representation is somatotopically organized: different extensions were found for the hands, trunk, head, and lower limbs ( ; ; ). It is modulated by several aspects of our everyday life, such as action ( ), danger and fear ( ), and stimuli valence ( ).

PPS is somatotopically modulated by the presence of SCI, but this representation can be recovered with an action-based training. While the upper limbs PPS representation in paraplegics is preserved ( ; ; ), the PPS around the lower limbs is shrunk ( ). However, applying passive mobilization on lower limbs can recover their PPS representation ( ).

The representation of the extrapersonal space is also modulated by action and metabolic energy. According to the Economy of Action principle ( ), the perceptual representation of the space around us is modulated by our metabolic energy, in order to spare energetic resources, mostly when they are lacking. Therefore, when we are in a low physiological condition, or in a physically demanding situation, distances to be (implicitly) walked are perceived longer, and slopes are perceived stepper ( ).

Indeed, experimental results showed that in estimating the slope of a hill or distance, people tend to overestimate when wearing a heavy backpack ( ). These effects may be elicited by poor physical health, fatigue ( ), while the opposite effect is appreciable after the assumption of energy beverages ( ). Thus, the characteristics of the perceptual representation of the extrapersonal space that are potentially linked to a bodily demanding action will be overestimated to prevent the action execution. Therefore, perceptual errors are natural, and they increase as the required effort increases.

In a virtual-reality task, participants affected by complete paraplegia saw a flag depicted on an ascending wheelchair ramp that could have different degrees of slant. Participants had to estimate the distance from the flag on their own wheelchair, or on a wheelchair never used before.

Distance estimation errors were modulated by the degrees of slant of the ramp only when they were seated on their own wheelchair, while on the other wheelchair, the estimation errors were constant ( ), suggesting that only when they are on their wheelchair they are able to implicitly represent the action of reaching the flag, allowing a representation of space that follows the Economy of Action principle ( ).

The effects of rehabilitation on the body, action, and space representations

Physiotherapy interventions can significantly affect, not only from a rehabilitative point of view, but they can likewise be beneficial for the body, action, and space representations.

In particular, the effects of physiotherapy training on body schema representation were studied on a sample of 21 individuals affected by SCI. Before starting the training, and at the end of it, participants underwent a Mental Body Rotation paradigm ( ), with stimuli depicting a left or right hand, foot, or a body ( ). Before the training, SCI participants and a group of healthy participants showed the same pattern for hands and bodies stimuli, while with feet stimuli, only the healthy group showed the bio-mechanical effect. Interestingly, the level of completeness of the lesion covaried with RTs with feet stimuli: slower RTs were associated with more complete lesions, suggesting a greater difficulty in processing stimuli concerning paralyzed body parts.

After the physiotherapy intervention, the bio-mechanical effect appeared also with feet stimuli in the SCI group, showing that a rehabilitative intervention has also effects on the body schema representation ( ).

Another fascinating study, added to the normal physiotherapy rehabilitation a motor imagery training to tetraplegic patients and a control group. Results from magneto-encephalography showed that before the training in the SCI group, there was a compensatory recruitment, with a higher activation of more areas during motor imagery tasks than the control group. Immediately after the training, and after 2 additional months, the SCI group showed a recruitment similar to the control group ( ).

The impact of rehabilitative training, in particular when they involve the mobilization of paralyzed limbs, is evident also in the PPS representation. In 2016, a study from our group noticed that PPS in complete paraplegics individuals is shrunk around the feet, though a passive mobilization lasting 15 min can recover it ( ).

Importantly, this recovery is not present when the passive-mobilization is only visual (e.g., the lower limbs are not actually moved, but participants observe a virtual-reality video showing their legs being moved, see Fig. 5 ).