Phantom learning: intermanual transfer of sequence learning in an amputee with phantom limb

Amputees who experience a phantom limb sometimes report that their phantom has certain sensory proprieties, like touch and pain, as well as kinesthetic proprieties, like being able to perform voluntary movements. Here, we focused on the motor domain and recruited one left upper-limb amputee (patient FC), who reported a vivid phantom limb and the ability to move it in a volitional manner. We asked whether the phantom movement go so far as leading to a motor learning, that, in turn, can be able to be transferred to the intact limb. To this aim, we took advantage from the intermanual transfer mechanism, that occurs when healthy subjects learn a motor skill with one hand and this results in performance improvement of the other hand as well. We tested patient FC and 10 aged-matched healthy controls, by using a sequence-learning task, in which the duration of the sequence execution was recorded with a sensor-engineered glove. The sequence duration was assumed as dependent variable to evaluate the ability to perform a fingers-thumb opposition sequence whit the right (intact) hand, before (naïve condition) and after a training with the left (phantom) hand. In the training phase, participants were asked either to actually execute the sequence (real condition) or to imagine it (imagery condition). The crucial aspect of this paradigm is that FC reported to be able to discriminate between the real and imagery training with her phantom. In healthy controls, results showed that, after a real training with the left hand, the ability to perform the sequence with the right hand was significantly improved with respect to the naïve condition (i.e., the sequence duration was significantly reduced). After the imagery training, no performance improvement was found (although the presence of a not significant tendency). Crucially, in FC, we found a significant performance improvement only after a “real” training with her phantom, suggesting the presence of an intermanual transfer. The first finding of the present study is that, in healthy subjects, an imagery training is not sufficient in order for the intermanual transfer to occur; a real motor learning seems to be necessary. In FC, we demonstrated that a) volitional movements with a phantom limb can induce an intermanual transfer comparable to that evident in real movements; b) motor execution and motor imagery with a phantom limb are functionally disentangled; c) neural mechanisms underpinning the intermanual transfer continue to operate despite the prolonged absence of any proprioceptive or visual feedbacks. Converging evidence show that phantom limb is not “imaginary”, but arises from physiological changes that occur after amputation. Here, we provided the first evidence that phantom limb can learn a motor skill and transfer it, through the callosal connections, to the intact limb.