The study by Cai et al. (2016) longitudinally investigated the relationship between cortical volume extracted from MRI and clinical outcome measures, both administered in the acute phase (within 5 days) and at 1 year follow-up from stroke onset. The authors recruited 11 acute stroke patients with ischaemic subcortical stroke. Cortical volume was estimated from T1-weighted images acquired on a 1.5 MR scanner. Voxel Based Morphometry (VBM) was performed with the VBM8 toolbox for SPM (https://www.fil.ion.ucl.ac.uk/spm/) to measure the potential changes in gray matter volume (GMV) after stroke, knowing that GMV contains information about CT and cortical surface area. The National Institutes of Health Stroke Scale (NIHSS) and Motricity Index (MI) were used as clinical outcome measures. Authors found a correlation between changes in GMV (corrected for lesion volume) and the clinical variables, specifically demonstrating that a more pronounced cortical atrophy in the precentral gyrus ipsilateral to stroke lesion correlated with a worse clinical recovery. Conversely, higher GMV of the contralateral orbitofrontal cortex (OFC) predicted a better motor recovery.

A recent work longitudinally investigated bilateral CT changes in basal ganglia stroke patients and healthy controls over a time-span of 6 months with 3 Tesla MRI (Liu et al., 2020). CT changes were assessed at five time points (within 7 days post-stroke and again at 14, 30, 90, and 180 days after the event) in 33 patients. Patients were divided into 2 groups according to whether or not the lesions affected the functional motor regions of the striatum, defined using resting state f-MRI (the striatal motor group – SMD and the non-striatal motor group – N-SMD). Patients' motor function was assessed by the Fugl-Meyer scale performed before and after each MRI scan. Structural MRI data were processed using Freesurfer. Fourteen patients were classified into the SMD and 19 into the N-SMD. The cortical thickness changes were explored comparing the baseline and 180 days post stroke images across all stroke participants and a significant increase in cortical thickness was found both in the ipsilesional (frontal pole, superior frontal gyrus, medial prefrontal cortex) and contralesional hemisphere (frontal pole, precentral gyrus, ventrolateral prefrontal cortex, superior frontal gyrus, medial prefrontal cortex, superior and middle temporal gyri). Furthermore, the evolution of these changes throughout the five time points was explored and a significant increase in CT over time was seen in the ipsilateral and contralateral hemisphere in the patient group; in comparison, healthy control participants demonstrated no cortical thickness changes over time or at any point. Moreover, the SMD and N-SMD groups underwent different patterns of cortical reorganization after stroke. The CT differences between 7 and 180 days post-stroke did not correlate with improvement in Fugl-Meyer scores suggesting that CT changes may not have a linear relationship with motor improvement. Instead, the SMD group exhibited a larger motor impairment compared to the N-SMD group, which underlines the importance of considering the stroke location when assessing symptoms and recovery. These study findings showed that CT changes appear over time and in cortical areas beyond the lesion site and that these changes are due to post stroke reorganization, given the fact that healthy controls did not show increase in CT. This increase in CT might be a product of motor recovery and motor compensation even if no correlation with motor score was found.

Cheng et al. (2020) reported another longitudinal study with a follow up of 1 year, conducted on 18 patients with subcortical ischaemic stroke and upper limb paresis. In this study the authors' aim was to test, if selective cortical atrophy of brain areas connected to subcortical stroke lesions was observable in the late chronic stage after stroke specifically in the contralesional homologous brain areas. T1-weighted and DTI images were acquired on a 3T MRI scanner and only NIHSS, FMA and grip strength were considered as clinical outcome measures. Clinical and imaging data were collected at 3–5 days after stroke and after 1 year. CT was measured using the Freesurfer software from T1-weighted images. Results showed that cortical atrophy involved regions connected with the lesion on both the affected and unaffected hemispheres. At the 1 year follow up no relationship between CT and clinical outcome was found.

The hypothesis that after a subcortical stroke there is a secondary decay of CT in the motor areas related to the degree of motor function impairment was longitudinally investigated in another study by Liu et al. (2015). With their work, they aimed at confirming that, after an initial reduction of CT, there is a progressive increase in the neuronal activity in motor areas, which might be correlated with motor function recovery after stroke. To confirm this hypothesis CT analysis from structural T1 MRI and a functional parameter from resting state MRI were combined. 3T MRIs were performed in the acute phase (before day 7 post stroke), after 4 weeks and after 12 weeks. The population included 22 patients with acute ischaemic subcortical stroke and 22 healthy controls. The authors assessed patients' motor function with FMA and NIHSS before each MRI examination and the correlation between motor outcome and changes in the investigated imaging indices was additionally explored. Authors found that CT was reduced in the premotor cortex, supplementary motor cortex (SMC) and precuneus 12 weeks after stroke. Interestingly, an increase was found in mean CT in the supplementary motor cortex and the insula of the unaffected hemisphere. The thickening of the contralesional SMC was correlated with changes in FMA scores and patients with significantly increased CT in the contralesional SMC experienced a greater improvement in motor function.

Furthermore, Kraemer et al. (2004) aimed at assessing the relationships between cortical volume in the acute and chronic phases after ischemic brain lesion and clinical recovery. Their study retrospectively included 10 patients affected from ischaemic stroke in the territory of the middle cerebral artery assessed with the European Stroke Scale (ESS), which was administered in combination with their MRIs. T1-weighted images were acquired on a 1.5T MRI scanner at acute and chronic stage, but time of acquisitions were not standardized, thus it should be considered as limitation. To assess cortical brain volumes they used MRI voxel-guided morphometry. Results revealed a shrinkage in cortical volume in brain areas exceeding the ischaemic lesion. Remote changes were found in the ipsilateral hemisphere and, in several patients, in the contralateral one. Brain volume alterations were not related to age, recovery or time between scans. The authors hypothesized that paresis and reduced usage may lead to secondary changes in cortical brain volume.

The post-stroke recovery variability issue drove the group by Yu et al. (2017) to investigate the differences in the gray matter volume and cerebral blood flow in acute vs. chronic subcortical stroke in depth. The background hypothesis was that comparing brain reorganization at different times from onset could provide insights into the functional and anatomical bases of recovery after stroke. In the study design the time interval between scans was variable (ranging from 3 to 8 months) and imaging results were correlated with BI and NIHSS as outcome measures. Data from 12 acute stroke patients were analyzed, high resolution T1-weighted MRIs were acquired on a 3T scanner and data were processed with the voxel-based morphometry 8 (VBM 8) toolbox for Statistical Parametric Mapping 8 (SPM8). Cortical volume was significantly reduced from acute to chronic phase in several regions of the ipsilesional hemisphere such as the post central gyrus, the precentral gyrus, the insula, the precuneus, the angular gyrus, the thalamus, and the anterior cerebellar lobe. Also, cortical thinning was found in the anterior and posterior cerebellar lobes. Cortical thickening was seen in the ipsilesional hippocampus and the contralesional orbital and the lingual gyrus. Only the anterior cerebellar lobe atrophy in the contralesional hemisphere was negatively correlated with recovery. Interestingly, atrophy in the precentral gyrus, the main area of voluntary movements, had no effects on stroke recovery. Based on these findings, the authors hypothesized that the increased volume in specific brain regions may be a compensatory response in promoting recovery after stroke.

Cheng et al. (2015) conducted a prospective study investigating the impact of subcortical stroke lesions on CT and on the recovery of upper limb function. Twelve acute ischaemic stroke patients with subcortical brain infarct were recruited and underwent a 3-month follow-up. The outcome of interest for functional recovery was the upper limb function, assessed by NIHSS, Action Research Arm Test (ARAT), Fugl-Mayer (FMA) and grip strength. Imaging data consisted of T1-weighted and DTI images acquired on a 3T MRI scanner combined with clinical examination (i.e., 3–5 days and 3 months after stroke). The study looked at CT changes between regions connected to the lesioned stroke area both in the affected and the unaffected hemisphere. Imaging data analysis was performed using the Freesurfer software package (Dale et al., 1999) (https://surfer.nmr.mgh.harvard.edu/). Clinical outcome measures of motor function improved in all patients. Results from cortical measurements showed significant cortical thinning involving the superior frontal gyrus and regions at the border of the supplementary motor area in the affected hemisphere, while non-significant changes were found in the unaffected hemisphere. The lesion size did not correlate with the CT changes. No direct effect of such thinning on upper limb motor performance was found and relative changes of clinical outcome measures were not significantly correlated with changes in CT. The authors speculated that the lack of correlation between CT changes and functional recovery could be due to the short follow-up period and the small sample size.

Gauthier et al. (2012), aimed at evaluating the relationships between chronic motor deficit of the upper limb in stroke patients and the amount of thinning in brain regions not apparently affected. In addition, the study aimed at addressing the question of whether regional gray matter thinning can be related to the improvement of upper limb function after Constraint Induced Movement Therapy (CIMT). To answer the research question, they recruited 85 chronic stroke patients and assessed their upper limb function with the MAL and WMFT. Forty-four subjects underwent MRI scans and T1-weighted MR images were acquired on a 1.5T or a 3T scanner; cortical features were measured with voxel-based morphometry (VBM). The authors investigated the relationships between gray matter density and upper limb recovery to calculate a pre-treatment motor status and subsequently looked at the same measures in the CIMT group. The authors found that better clinical outcome, and in this case also clinical benefit after CIMT, was related to cortical thickness of the sensory-motor regions of the healthy hemisphere.

The relationship between CIMT and cortical morphology was also investigated by Sterr et al. (2013). The authors looked at CT variations in 31 patients with moderate to severe chronic stroke sequelae and compared CIMT vs. standard therapy. A 3T MRI scanner was used to acquire T1-weighted MR images that were processed for surface based analysis with Freesurfer. To assess motor function, the MAL and WMFT were administered before and after therapy. Results showed a significant improvement in all the clinical measures in the CIMT when compared with the control group. CT changed in the precentral and the post central gyrus of the non-lesioned hemisphere, no difference between groups was found. Furthermore, no significant cortical changes associated with modification in clinical outcomes were seen. The authors speculated that the variation of cortical properties reflected the use-dependent structural changes, such as the increase in synapses, dendrites, axonal spines and glial cells.

Focusing on chronic patients, Buetefisch et al. (2018) aimed at evaluating, if impairment of hand function is related to structural and functional reorganization of the primary motor cortex (M1) and its corticospinal projections of the lesioned hemisphere. Eighteen patients with cortical and subcortical ischaemic infarction involving the primary motor area and the corticospinal tract (CST) were studied. Hand motor function was assessed with the Jebsen-Taylor test and a kinematic measure of hand function, i.e., the peak acceleration of wrist extension movements. Furthermore, the Wolf Motor Function Test (WMFT) and the Motor Activity Log (MAL) were used. Data from patients were compared with data from two groups of age-matched healthy subjects. It was found that the primary motor cortex of the affected hemisphere was thinner than the contralateral. A significant association between hand function and structural integrity of the primary motor system, as measured by the primary motor cortex thickness and cortical anisotropy of the cortico-spinal tract, was not found. A limitation of this study is that it was not possible to distinguish if the process leading to the changes in M1 CT is regenerative or degenerative since they didn't have a second set of measurements for comparison.

A definite focus on the neural substrate underpinning upper limb recovery after stroke was the aim of Jones et al. (2016) who used biochemical and MRI approaches within the framework of a case control study design. A total of 17 patients with chronic subcortical stroke and 11 healthy controls were recruited. The authors wanted to quantify anatomical and metabolic differences in the primary motor cortex and to look at their relationships with the hemiparetic arm function recovery in chronic stroke patients. Participants underwent 3T MRI scans and functional assessment. Cortical reconstruction and segmentation were performed with the Freesurfer software. Total N-acetylaspartate (t-NAA) and glutamate (Glx) concentrations were measured, as both are altered in chronic stroke and their level of change is related to arm impairment and CT (Cirstea et al., 2011, 2012). Upper limb motor function was assessed with the WMFT. A significant positive correlation was found between tNAA and M1 thickness for ipsilesional and contralesional hemispheres in the stroke group, and tNAA concentration explained a larger amount of variance in motor performance. The precentral gyrus thickness was significantly lower in the stroke group where ipsilesional thickness was not significantly associated with motor performance.

Finally, Chen et al. (2019) highlighted the somatosensory deficits in thalamic stroke with the aim to identify cortical regions causally influenced by the ischaemic damage and to determine the association between structural/functional alteration and somatosensory impairment. To fulfill this purpose, 31 patients with chronic sensory-motor impairments secondary to thalamic infarction and 32 age and sex-matched healthy controls were enrolled. Clinical examination included NIHSS, FMA, Barthel Index and Lindmark assessment. MRIs were acquired with a 3T scanner and cortical measures were extracted from the T1-weighted images employing Freesurfer. Results revealed decreased cortical volume in the ipsilesional primary somatosensory cortex demonstrating a link between alteration of the cortical volume and somatosensory impairment.