Electrophysiological validation of diffusion-tensor tractography; evidence from pediatric deep brain stimulation

Abstract: Diffusion tensor imaging (DTI) variables including diffusivity, anisotropy, and fiber tract length are thought to reflect functional physiological properties of brain connectivity.  In our hospital, we perform individualized targeting for deep brain stimulation (DBS) in children using temporary depth electrodes implanted in potential targets, with stimulation and recording over 1 week.  As part of this testing, we are able to determine the amplitude and latency of evoked potentials reflecting the connectivity between deep brain stimulation sites and recording sites.  We are also able to determine the non-stimulated relationship between activity at different areas using a transfer function analysis analogous to bandlimited coherence.  By comparing the electrical evoked responses and transfer function power to DTI variables we show that, as expected, amplitude of response correlates with number and size of fibers, while delay correlates with fiber diameter.  Surprisingly, in our dataset the delay does not correlate with fiber length, which is explained by a compensatory increase in fiber diameter for larger fiber lengths.  This suggests that in some regions, larger fibers are used to carry signals over longer distances which may contribute to synchronous arrival of signals at both near and distant targets.