Final Report Summary - PHANTOMMIND (Phantom phenomena: A window to the mind and the brain)
Phantom experiences occur in almost all amputees but are among the least understood sensory phenomena. Recently, changes in the representation of body maps in the brain were found to be related to phantom pain and it has also been demonstrated that there are great similarities between non-painful phantoms and bodily illusions such as the rubber hand illusion. This research has also shown that the brain does not process the physical but the perceived reality, which opens the door to manipulations of the perceived reality in basic research and the treatment of phantom pain. Behavioural intervention methods such as prosthesis, sensory discrimination or mirror training influence phantom limb pain and alter brain function. Thus, phantom phenomena are an excellent tool to study the neural basis of somatosensory and specifically bodily perception and this may lead to new treatment methods such as brain-computer interfaces or virtual reality applications for phantom pain and similar pain states.
The aim of this proposal was (WP1) an exact assessment and description as well analysis of the interrelationship of painful and nonpainful phantom phenomena such as phantom limb awareness, nonpainful phantom sensation, phantom limb pain, telescoping (the shrinking or expansion of the phantom limb), phantom limb position and movement, stump-related sensations and pain, prosthesis use, empathy, bodily awareness and proneness to bodily as well as other sensory illusions in a large sample of amputees, (WP2) the analysis of the neural underpinnings of these phenomena in small selected subgroups of amputees, using functional magnetic resonance imaging, (WP3) the analysis of the person-related determinants and the neural correlates of bodily illusions in a sample of healthy controls in an effort to determine potential common neural correlates between phantoms, abnormal body sensations and illusionary percepts and (WP4) use of virtual reality training after amputation in order to determine to what extent manipulations of the body image and sensorimotor and visual feedback alter the development and the brain correlates of phantom phenomena.
We have built a data base consisting of the questionnaire data from more than 3.600 unilateral amputees of our nation-wide study. To reach this numbers more than 40.000 people received letters with the questionnaires. The interrelationship of the questionnaire data was assessed with structural equitation modelling to specify which variables determine phantom limp pain. Variance was declared by pain variables as the intensity of residual limb pain and pain prior to amputation, phantom sensations variables as the intensity of phantom sensations and the intensity of telescoping and demographic and clinical variables as age at amputation and length of the residual limb. Further on the relationship of phantom pain and dreaming was assessed with regression analysis (Bekrater-Bodmann et al., accepted with minor revisions).
To analyze the neural underpinnings of the painful and nonpainful phantom phenomena more than 345 amputees with and without phantom limb pain were tested in our structural and functional imaging studies. The basis for these experiments was a study of our group (Diers et al., 2010), showing that amputees with and without phantom limb pain differ in their neural activity in brain areas representing the missing limb. We developed a virtual reality application of the mirror box illusion (Diers et al., 2015) and the rubber hand illusion (Bekrater-Bodmann et al., 2014) which is applicable in the high magnetic environment of a magnetic resonance imaging (MRI) scanner. In the mirror box illusion participants move one hand (the intact hand in amputees) and perceiving the mirrored hand as the movement of the non-moving (or amputated) hand. During the rubber hand illusion participants perceive a rubber hand as belonging to their body by touching the hidden hand or residual limb with synchronous visually feedback of such touching on the virtual hand. Using this setup in a sample of upper-limb amputees, we showed that even amputees can perceive an artificial limb as belonging to their body. These findings indicate that the part of SI representing the phantom limb can be activated through multisensory stimulation, which might have important implications for the design of prostheses.
To assess the neural correlates of referred sensations (RS, sensations outside of a stimulated area triggered by the stimulation e.g. perceiving sensations on the phantom hand in unilateral arm-amputees by stimulating the mouth area) we identified within a neuropsychological examination body areas triggering RS in the phantom limb. Afterwards participants were measured in the MRI-scanner during stimulation of this body area triggering the RS as well as a control body side. Preliminary fMRI-results indicate a re-activation of the somatosensory representation of the amputated limb (phantom cortex) during the perception of RS. This result suggests that RS might operate by mechanisms comparable to the visual mirror feedback i.e. by re-activating the maladaptively reorganized phantom cortex.
In addition, we induced the mirror box illusion under congruent and incongruent (asynchronous tactile and visual feedback) feedback conditions in a large sample of more than 200 healthy participants. Our results indicate that asynchronous stimulation (incongruent condition) not only dramatically diminishes the vividness of the mirror illusion, but induce complex new sensations such as somatic misperceptions and the perception of owning three limbs (Foell et al., 2013). This finding has important implications for mirror treatment. Mirror treatment is effective in reducing phantom limb pain and is accompanied with changes in the somatosensory cortex (Foell et al., 2014). We could show that the induction of ownership for the mirrored limb is crucial for the effectiveness of mirror treatment. However, perceptual alterations related to the phantom limb (i.e. telescopic distortions) were identified as strong predictors of ineffectiveness of the treatment, highlighting the importance of multisensory congruence and higher-order body perceptions for this kind of treatment. To advance the mirror box approach, we developed an augmented reality training system (Trojan, Diers et al., 2014). In this training several tasks were implemented to make the training more exciting and increase the commitment of the patients. A head-mounted-display equipped with cameras captures one hand (the intact hand in amputees) held in front of the body, mirrors it, and displays it in real-time in (Bach et al. 2010). The patient’s neuronal responses to the rubber hand illusion and the mirror box illusion our paradigms were assessed prior to and directly after the training phase. These virtual reality applications are promising and could be extended in the future.
In addition to the publications listed in the annex we still have 4 additional publications under review and preparing 6 additional manuscripts which will be published in the course of the year.
The aim of this proposal was (WP1) an exact assessment and description as well analysis of the interrelationship of painful and nonpainful phantom phenomena such as phantom limb awareness, nonpainful phantom sensation, phantom limb pain, telescoping (the shrinking or expansion of the phantom limb), phantom limb position and movement, stump-related sensations and pain, prosthesis use, empathy, bodily awareness and proneness to bodily as well as other sensory illusions in a large sample of amputees, (WP2) the analysis of the neural underpinnings of these phenomena in small selected subgroups of amputees, using functional magnetic resonance imaging, (WP3) the analysis of the person-related determinants and the neural correlates of bodily illusions in a sample of healthy controls in an effort to determine potential common neural correlates between phantoms, abnormal body sensations and illusionary percepts and (WP4) use of virtual reality training after amputation in order to determine to what extent manipulations of the body image and sensorimotor and visual feedback alter the development and the brain correlates of phantom phenomena.
We have built a data base consisting of the questionnaire data from more than 3.600 unilateral amputees of our nation-wide study. To reach this numbers more than 40.000 people received letters with the questionnaires. The interrelationship of the questionnaire data was assessed with structural equitation modelling to specify which variables determine phantom limp pain. Variance was declared by pain variables as the intensity of residual limb pain and pain prior to amputation, phantom sensations variables as the intensity of phantom sensations and the intensity of telescoping and demographic and clinical variables as age at amputation and length of the residual limb. Further on the relationship of phantom pain and dreaming was assessed with regression analysis (Bekrater-Bodmann et al., accepted with minor revisions).
To analyze the neural underpinnings of the painful and nonpainful phantom phenomena more than 345 amputees with and without phantom limb pain were tested in our structural and functional imaging studies. The basis for these experiments was a study of our group (Diers et al., 2010), showing that amputees with and without phantom limb pain differ in their neural activity in brain areas representing the missing limb. We developed a virtual reality application of the mirror box illusion (Diers et al., 2015) and the rubber hand illusion (Bekrater-Bodmann et al., 2014) which is applicable in the high magnetic environment of a magnetic resonance imaging (MRI) scanner. In the mirror box illusion participants move one hand (the intact hand in amputees) and perceiving the mirrored hand as the movement of the non-moving (or amputated) hand. During the rubber hand illusion participants perceive a rubber hand as belonging to their body by touching the hidden hand or residual limb with synchronous visually feedback of such touching on the virtual hand. Using this setup in a sample of upper-limb amputees, we showed that even amputees can perceive an artificial limb as belonging to their body. These findings indicate that the part of SI representing the phantom limb can be activated through multisensory stimulation, which might have important implications for the design of prostheses.
To assess the neural correlates of referred sensations (RS, sensations outside of a stimulated area triggered by the stimulation e.g. perceiving sensations on the phantom hand in unilateral arm-amputees by stimulating the mouth area) we identified within a neuropsychological examination body areas triggering RS in the phantom limb. Afterwards participants were measured in the MRI-scanner during stimulation of this body area triggering the RS as well as a control body side. Preliminary fMRI-results indicate a re-activation of the somatosensory representation of the amputated limb (phantom cortex) during the perception of RS. This result suggests that RS might operate by mechanisms comparable to the visual mirror feedback i.e. by re-activating the maladaptively reorganized phantom cortex.
In addition, we induced the mirror box illusion under congruent and incongruent (asynchronous tactile and visual feedback) feedback conditions in a large sample of more than 200 healthy participants. Our results indicate that asynchronous stimulation (incongruent condition) not only dramatically diminishes the vividness of the mirror illusion, but induce complex new sensations such as somatic misperceptions and the perception of owning three limbs (Foell et al., 2013). This finding has important implications for mirror treatment. Mirror treatment is effective in reducing phantom limb pain and is accompanied with changes in the somatosensory cortex (Foell et al., 2014). We could show that the induction of ownership for the mirrored limb is crucial for the effectiveness of mirror treatment. However, perceptual alterations related to the phantom limb (i.e. telescopic distortions) were identified as strong predictors of ineffectiveness of the treatment, highlighting the importance of multisensory congruence and higher-order body perceptions for this kind of treatment. To advance the mirror box approach, we developed an augmented reality training system (Trojan, Diers et al., 2014). In this training several tasks were implemented to make the training more exciting and increase the commitment of the patients. A head-mounted-display equipped with cameras captures one hand (the intact hand in amputees) held in front of the body, mirrors it, and displays it in real-time in (Bach et al. 2010). The patient’s neuronal responses to the rubber hand illusion and the mirror box illusion our paradigms were assessed prior to and directly after the training phase. These virtual reality applications are promising and could be extended in the future.
In addition to the publications listed in the annex we still have 4 additional publications under review and preparing 6 additional manuscripts which will be published in the course of the year.