Table 2.

Summary of randomized controlled trials in assessing brain changes related to cognitive rehabilitation programs in PD.

Authors	MRI sample	H&Y	Cognitive rehabilitation program	MRI acquisition/preprocessing/analysis	MRI statistical analysis	MRI results	Correlation with cognitive measures	Limitations/risk of bias
Cerasa et al. [33]	15 PD 8 CR 7 ACG	1–3	RehaCom computer program Training: (i) Attention (ii) Information processing	Resting-state fMRI/ICA in FSL/FC analysis	ANOVA (group × time) Region of interests Dorsolateral PFC Ventrolateral PFC ACC Sup + inf parietal left Caudate Cerebellum	Increased functional activity: (i) Left dorsolateral PFC (executive network) (ii) Superior parietal left (attention network)	—	Small sample size Only one type of MRI acquisition
Díez-Cirarda et al. [38] (cognitive results in [31])	30 PD 15 CR 15 ACG	1–3	“REHACOP” program Training: (i) Attention (ii) Memory (iii) Language (iv) EF (v) Social cognition	Resting-state fMRI/ROI-to-ROI approach in CONN toolbox	ANOVA (group × time) Paired t-test Whole brain	Increased FC between BA9L-BA20L and BA9R-BA20L	Yes	Small sample size and reduced at long-term follow-up Memory fMRI paradigm results at FWE uncorrected
				Memory fMRI paradigm/model-based approach in SPM		Increased brain activation Learning task: left frontal inferior lobe
						Increased brain activation Memory task: left middle temporal lobe
				Diffusion weighted/TBSS in FSL		—
				T1 weighted/VBM in FSL		—

ACC = anterior cingulate cortex; ACG = active control group; BA = Brodmann area; CR = cognitive rehabilitation; EF = executive functions; FC = functional connectivity; fMRI = functional magnetic resonance imaging; FSL = FMRIB Software Library; H&Y = Hoehn and Yahr; ICA = independent component analysis; MRI = magnetic resonance image; PD = Parkinson's disease; PFC = prefrontal cortex; TBSS = tract-based spatial statistics; VBM = voxel-based morphometry.