Memory loss is one of the first symptoms of typical Alzheimer's disease (AD), for which there are no effective therapies available. The precuneus (PC) has been recently emphasized as a key area for the memory impairment observed in early AD, likely due to disconnection mechanisms within large-scale networks such as the default mode network (DMN). Using a multimodal approach we investigated in a two-week, randomized, sham-controlled, double-blinded trial the effects of high-frequency repetitive transcranial magnetic stimulation (rTMS) of the PC on cognition, as measured by the Alzheimer Disease Cooperative Study Preclinical Alzheimer Cognitive Composite in 14 patients with early AD (7 females). TMS combined with electroencephalography (TMS-EEG) was used to detect changes in brain connectivity. We found that rTMS of the PC induced a selective improvement in episodic memory, but not in other cognitive domains. Analysis of TMS-EEG signal revealed an increase of neural activity in patients' PC, an enhancement of brain oscillations in the beta band and a modification of functional connections between the PC and medial frontal areas within the DMN. Our findings show that high-frequency rTMS of the PC is a promising, non-invasive treatment for memory dysfunction in patients at early stages of AD. This clinical improvement is accompanied by modulation of brain connectivity, consistently with the pathophysiological model of brain disconnection in AD.

This study indicates that disconnection precedes GM atrophy in the PCC, which is a critical area of the DMN, and supports the hypothesis that GM atrophy in specific regions of AD brains likely reflects a long term effect of brain disconnection. In this context, our study indicates that GM atrophy in PCC accompanies the conversion from MCI to AD.

Granulin (GRN) mutations have been identified as a major cause of frontotemporal lobar degeneration (FTLD) by haploinsufficiency mechanism, although their effects on brain tissue dysfunction and damage still remain to be clarified. In this study, we investigated the pattern of neuroimaging abnormalities in FTLD patients, carriers and noncarriers of GRN Thr272fs mutation, and in presymptomatic carriers. We assessed regional gray matter (GM) atrophy, and resting (RS)-functional magnetic resonance imaging (fMRI). The functional connectivity maps of the salience (SN) and the default mode (DMN) networks were considered. Frontotemporal gray matter atrophy was found in all FTLD patients (more remarkably in those GRN Thr272fs carriers), but not in presymptomatic carriers. Functional connectivity within the SN was reduced in all FTLD patients (again more remarkably in those mutation carriers), while it was enhanced in the DMN. Conversely, presymptomatic carriers showed increased connectivity in the SN, with no changes in the DMN. Our findings suggest that compensatory mechanisms of brain plasticity are present in GRN-related FTLD, but with different patterns at a preclinical and symptomatic disease stage.

This study investigates abnormalities of grey (GM) and white matter (WM) in Alzheimer's disease (AD), by modeling the AD pathological process as a continuous course between normal aging and fully developed dementia, with amnesic mild cognitive impairment (aMCI) as an intermediate stage. All subjects (9 AD, 16 aMCI patients, and 13 healthy controls) underwent a full neuropsychological assessment and an MRI examination at 3 Tesla, including a volumetric scan and diffusion tensor (DT)-MRI. The volumes were processed to perform a voxel-based morphometric analysis of GM and WM volume, while DT-MRI data were analyzed using tract based spatial statistics, to estimate changes in fractional anisotropy and mean diffusivity data. GM and WM volume and mean diffusivity and fractional anisotropy were compared across the three groups, and their correlation with cognitive functions was investigated. While AD presented a pattern of widespread GM atrophy, tissue loss was more subtle in patients with aMCI. WM atrophy was mainly located in the temporal lobe, but evidence of WM microscopic damage, assessed by DT-MRI, was also observable in the thalamic radiations and in the corpus callosum. Memory and executive functions correlated with either GM volume or fractional anisotropy in fronto-temporal areas. In conclusion, this study shows a comprehensive assessment of the brain tissue damage across AD evolution, providing insights on different pathophysiological mechanisms (GM atrophy, Wallerian degeneration, and brain disconnection) and their possible association with clinical aspects of cognitive decline.

Multivariate pattern analysis and statistical machine learning techniques are attracting increasing interest from the neuroimaging community. Researchers and clinicians are also increasingly interested in the study of functional-connectivity patterns of brains at rest and how these relations might change in conditions like Alzheimer's disease or clinical depression. In this study we investigate the efficacy of a specific multivariate statistical machine learning technique to perform patient stratification from functional-connectivity patterns of brains at rest. Whilst the majority of previous approaches to this problem have employed support vector machines (SVMs) we investigate the performance of Bayesian Gaussian process logistic regression (GP-LR) models with linear and non-linear covariance functions. GP-LR models can be interpreted as a Bayesian probabilistic analogue to kernel SVM classifiers. However, GP-LR methods confer a number of benefits over kernel SVMs. Whilst SVMs only return a binary class label prediction, GP-LR, being a probabilistic model, provides a principled estimate of the probability of class membership. Class probability estimates are a measure of the confidence the model has in its predictions, such a confidence score may be extremely useful in the clinical setting. Additionally, if miss-classification costs are not symmetric, thresholds can be set to achieve either strong specificity or sensitivity scores. Since GP-LR models are Bayesian, computationally expensive cross-validation hyper-parameter grid-search methods can be avoided. We apply these methods to a sample of 77 subjects; 27 with a diagnosis of probable AD, 50 with a diagnosis of a-MCI and a control sample of 39. All subjects underwent a MRI examination at 3T to obtain a 7minute and 20second resting state scan. Our results support the hypothesis that GP-LR models can be effective at performing patient stratification: the implemented model achieves 75% accuracy disambiguating healthy subjects from subjects with amnesic mild cognitive impairment and 97% accuracy disambiguating amnesic mild cognitive impairment subjects from those with Alzheimer's disease, accuracies are estimated using a held-out test set. Both results are significant at the 1% level.