Training-related increase of cortico-cerebellar connectivity in cerebellar degeneration patients as a function of feedback

Abstract

Cerebellar patients can improve motor performance with practice. However, it is unclear which type of feedback during practice is most beneficial to them. In this longitudinal intervention study, we investigated the effects of a five-day forearm movement training on cortico-cerebellar connectivity in patients with cerebellar degeneration and age- and sex-matched healthy controls (N=80, aged 55.4 ± 11.1 years). Participants were pseudo randomly assigned to one of four motor training conditions, varying online visual feedback (Vision/No Vision) and post-movement verbal feedback (Expl. Feedback / No Expl. Feedback). Anatomical and resting-state fMRI (rs-fMRI) was collected on the days before and after training. Behavioral and VBM data have been reported (Draganova et al. 2022). First, we derived a study-specific template from 40 anatomical scans, balanced across groups, timepoints and conditions. Compared to existing templates, the study template led to strongly improved alignment of individual fissures, reducing their spatial spread by 43% (cross-validated using the 40 held-out anatomical scans; t(38)= -18.06, p<0.0001). Second, to remove structured noise from rs-fMRI, we hand-labelled noise components derived from single-subject independent component analysis (ICA) to train an automated classifier FIX to identify noise components. Leave-one-out classification accuracy for noise classification was 50% higher when normalizing scans to MNI space via our study template, compared to direct normalization as implemented in FIX (t(42)=8.54, p<0.0001). Third, we defined regions of interests (ROIs) in neocortex and cerebellum, including dorsal premotor cortex (PMd), posterior parietal cortex (PPC) and cerebellar motor regions (region 1 and 2 in King et al., 2019). We quantified connectivity as Pearson’s correlations between the average BOLD time course extracted from the ROIs. At baseline, patients showed impaired cortico-cerebellar connectivity (χ2(2)= 16.82, p<0.001). Training with vision and explicit post-movement feedback (Vision + Expl. Feedback) increased left PMd – right cerebellar connectivity for all participants (χ2(3) = 9.01, p = 0.03). After training with vision and explicit feedback, patients exhibited an increase in right PMd – left cerebellar connectivity (i.e., contralateral to the trained arm; χ2(3)= 9.47, p = 0.02), such that connectivity after training was not different from controls (F(1,133) = 0.028, p = 0.87). Training with vision increased PPC –cerebellar connectivity contra- and bilaterally to the trained arm (Contralateral χ2(1) = 7.42, p < 0.01; Ipsilateral χ2(1) = 5.01, p = 0.03). We found no significant effect of explicit feedback on right dlPFC – left cerebellar Crus 2 connectivity. Our results show that a) cerebellar patients show impaired cortico-cerebellar connectivity, b) motor training increases cortico-cerebellar connectivity and c) connectivity in patients seems to benefit most from visuomotor training with explicit feedback.

Caroline Nettekoven

Postdoctoral Researcher

I am interested in the neural basis of complex behaviour. To study this, I use neuroimaging techniques, computational modelling of behaviour and brain stimulation.