Skip to main content
. 2023 May 17;14:1144228. doi: 10.3389/fneur.2023.1144228

Figure 3.

Figure 3

Lesion network-symptom mapping for depressive symptoms 6 months post-stroke and network damage scores. Results from the LNSM with the continuous HADS-D scale at 6 months post-stroke. (A) The unthresholded map of T-values shows a bilateral frontal, temporal and basal ganglia maximum, but there was no significant association between functional lesion network map strength and depressive symptoms [p(FWE) < 0.05 at cluster-level]. The analysis remained negative also when applying a mask for the bilateral middle frontal gyrus, when using a binary cut-off at HADS-D > 10 and when uncorrected for multiple comparisons (not shown here). (B) The depression circuit was derived from a region-of-interest (ROI) with a 9 mm diameter sphere (shown in red) around the peak coordinates (x = −32, y = 12, z = 36) reported by Padmanabhan and colleagues. The ROIs whole brain functional connectivity was calculated with the normative connectome of 100 healthy subjects. Warmer colors indicating positive connectivity to the ROI and cool colors negative connectivity. This constitutes the ‘depression circuit’ as described by Padmanabhan and colleagues. In green six random lesions of patients with and without depressive symptoms (binary cut-off > 10 on the HADS-D) are shown for illustration as an overlay on the ‘depression circuit’ – the intersection of the lesions with the depression circuit then results in the network damage score. (C) The network damage scores of patients with and without severe depressive symptoms did not differ significantly (p = 0.93). All data points are shown in the box plot with the exception of one outlier (NDS 6.3 × 104, depression group).