Fig. 6: Resected parcels in seizure-free patients show a lower weighted degree and higher modal controllability than equivalent parcels in controls and compared to virtual resections in patients.
a–c Changes in the mean Z scores for weighted degree, average controllability and modal controllability for each resective surgery patient between baseline, the a resected parcels, b the non-resected parcels and c the post-operative connectome. There were significant differences in the mean weighted degree (p = 0.002 and p = 6 × 10−4) and modal controllability (p = 6 × 10−4 and p = 4 × 10−4) for both resected and non-resected parcels in seizure-free patients only (paired t tests, adjusted for multiple comparisons using the Benjamini–Hochberg FDR method). For the post-operative connectome, there were significant differences in the mean weighted degree for both seizure-free and not seizure-free patients (p = 5 × 10−4 and p = 3 × 10−5, respectively), mean average controllability in seizure-free patients only (p = 0.001) and mean modal controllability for both seizure-free and not seizure-free patients (p = 3 × 10−6 and p = 4 × 10−5, respectively) (paired t tests, adjusted for multiple comparisons using the Benjamini–Hochberg FDR method). d Z scores of weighted degree, average controllability and modal controllability for the real resection compared to 1000 virtual resections for each of the patients, stratified by post-operative outcome. Bars show mean ±95% confidence intervals. The real resection parcels were significantly different from 0 for weighted degree (p = 0.007) and modal controllability (p = 0.002) in seizure-free patients only (one-sample t tests, adjusted for multiple comparisons using the Benjamini–Hochberg FDR method).
