Introduction: Functional networks in resting-state fMRI are identified by characteristics of their intrinsic low-frequency oscillations, more specifically in terms of their synchronicity. In advanced aging and clinical populations, this synchronicity among functionally linked regions is known to decrease, which may be associated with the observed cognitive and behavioral changes. Previous work from our group has revealed oscillations at the slow-5 range (0.01-0.027 Hz) are particularly susceptible to disruptions in aging and stroke.
Hypothesis: In this study, we aim to characterize longitudinally slow-5 oscillations in stroke patients for hope to observe a recovery of those impaired oscillations in the later stage of the stroke.
Methods: We followed a group of ischemic stroke patients (n=20) and healthy older adults (n=14) over two visits separated by a minimum period of three months. For the stroke patients, one visit occurred in their subacute window (10 days to 6 months after stroke), the other happened over their chronic window (> 6 months after stroke). Using a groupICA on 10-minutes eyes-close resting-state fMRI data, we assessed the frequency distributions of component’s representative time-courses for differences between the two time-points, in particular in regards to slow-5 spectral power.
Results: Slow-5 oscillations exhibited lower amplitudes in the stroke patient in their subacute stage in comparison to their healthy counterparts as was presented previously. Over time in their chronic stage, those patients show a recovery of those oscillations (p<0.005), reaching near equivalence to the healthy older adult group.
Conclusions: Our results indicate disrupted network dynamics in the subacute stroke population and demonstrate that return to normal level of spectral power by the chronic stage, potentially marking a recovery of the cortical system. This also verifies the implication of slow-5 oscillations in the integrity of network function in normal, but also in injured cortical systems.