144 CHAPTER 7 conclude that the reduced activity in these visuomotor areas during imagined movement of the affected upper extremity likely points to a deficit in the processing of self-relevant information and sensorimotor integration in patients with NA. Linking back to the bigger picture, we can speculate about how we interpret the interplay between these cerebral adaptations, peripheral nerve damage and persistent scapular dyskinesia. Because of peripheral nerve damage and subsequent motor weakness in NA, the movement a patient plans to make is different from the movement that is actually executed. As a result, the predicted sensory outcomes (i.e. efference copy), and actual sensory outcomes (i.e. action feedback) of the movement do not align. The fact that the sensory input, across modalities, will be different than expected, may lead to a sensory mismatch. 216, 217 Altered afferent feedback related to nerve damage may further contribute to this mismatch between the efference copy and the actual action feedback. This mismatch may lead to the problems with integration of visual and proprioceptive information about the upper extremity in the context of motor planning we find here. Once the peripheral nerves have (largely) recovered and the patient should once again be capable of healthy movement, the initial adaptations in visuomotor processing may continue to hinder motor planning and execution. Taken together, we can conclude that cerebral adaptations (also) occur in visuomotor areas, outside the core sensorimotor system, and that these changes may play a role in the persistence of motor dysfunction despite peripheral nerve recovery in NA. Clinical relevance of cerebral alterations and improving rehabilitation after peripheral nerve damage in NA We ended the previous paragraph with a proposed interplay between altered cerebral processes, peripheral nerve damage and clinical outcome in NA. A natural, and pivotal, follow-up question pertains to the clinical relevance of cerebral adaptations for recovery and rehabilitation in NA. Despite the fact that specific rehabilitation, targeting the central motor system, elicited greater improvement in functional capability of the upper extremity compared to usual care, we could not find corresponding group differences in the central nervous system. We also did not find longitudinal correlations between brain activity, task performance and clinical outcome. This could call the clinical relevance of the cerebral adaptations into question. Importantly though, we did find cross-sectional correlations across the three domains, where patients who had more difficulty with imagined movement with their affected arm, also had reduced activity during those movements and reported more pain. Moreover, despite the lack of longitudinal correlations between clinical outcome, brain activity and task performance, all three domains did show changes in the same direction: NA patients did not only have increased functional capability and reduced pain after treatment, they also showed increased brain activity in the direction of normality and were more accurate when imagining complex movements with their affected upper extremity at follow-up compared to baseline. These facts indicate that there is indeed a connection between cerebral adaptations, task performance, and clinical outcome in NA. Having established this connection, we can look at how our findings can be used to further improve rehabilitation for these patients. To improve rehabilitation, we need to know what processes and structures to target, when to intervene, which patients to select, and finally how to treat them. 218, 219 In the following paragraphs I will take a closer look at our findings to identify directions that could help determine those