80 CHAPTER 4 Group differences We observed an interaction between GROUP, LATERALITY, and BIOMECHANICAL COMPLEXITY. Specifically, when contrasting the effect of biomechanical complexity of right versus left hands, NA patients showed less brain activity than healthy participants in two clusters: (1) an area in the right extrastriate cortex, located at the junction of the occipital, parietal and temporal lobes, just anterosuperior to the extrastriate body area (EBA), covering parietal area G, posterior (PGp) in the inferior parietal lobule (18%), and part of V5 (12%); and (2) an area extending bilaterally along the parieto-occipital sulcus, covering parts of Brodmann area (BA) 17 (5% left, 6% right) and 18 (10% left, 8% right) (Figure 4A, Figure 4C-D, Table 3). The right extrastriate area falls inside a larger region where brain activity is sensitive to the biomechanical complexity of right hand-stimuli (Figure 4B, Supplementary Table 1). Post-hoc, we tested for a difference between right (affected) versus left (unaffected) hand stimuli on trials associated with biomechanically complex movements only. This revealed a group difference in the same areas as outlined above: compared tocontrols, NApatients showed reducedbrainactivity in right extrastriate cortex (PGp: 21%; V5: 10%) and a cluster along the bilateral parieto-occipital sulcus (BA17: 7% left, 5% right; BA18: 7% left, 7% right), as well as a small cluster in the superior parietal cortex (7A: 5% left, 5% right), specifically during imagery of biomechanically complex movements involving their affected limb (Figure 4A, Figure 4D, Table 3). There were no group differences for the main factors LATERALITY, BIOMECHANICAL COMPLEXITY and POSTURE, or for the interaction between LATERALITY and POSTURE. Post-hoc functional connectivity analysis Across both groups, the right extrastriate seed was functionally connected to several brain areas that were sensitive to biomechanical complexity, including cerebellum, premotor and primary motor cortex, frontal gyri and inferior/superior parietal cortex (Supplementary Figure 1A, Supplementary Table 3). Across groups, the bilateral parietooccipital sulcus was functionally connected to regions that together form the default mode network, as well as regions not typically part of the default mode network such as pre- and postcentral gyri (see Supplementary Figure 1B, Supplementary Table 3). There were no functional connectivity differences between groups. Brain-behaviour-symptom correlations NA patients with more persistent pain had significantly less activity in the right extrastriate cortex related to complex movements of the affected limb (r = -0.45, p = 0.004; Figure 5A). Other measures of symptom severity showed a similar tendency: patients with less activity in the right extrastriate cortex tended to have lower functional capability of the upper extremity (r =-0.30, p = 0.066), and a lower relative serratus anterior muscle strength on their affected side (r = 0.28, p = 0.083). Patients with more pain also had significantly greater difficulty with complex movements of their affected limb (i.e. greater positive difference in RT between affected and unaffected limb, matching the contrast for which patients had decreased cerebral activity in extrastriate cortex and parieto-occipital sulcus) (r = 0.33, p = 0.04; Figure 5C) and tended to have higher overall RTs (r = 0.29, p = 0.07). Moreover, patients with greater difficulty with complex movements of their affected limb had significantly less activity along the parieto-occipital sulcus (r =-0.42, p = 0.008; Figure 5B), but not in the right extrastriate cortex (r = -0.08, p = 0.62), when imagining