Cerebellar involvement in cognition, as well as in sensorimotor control, is increasingly recognized and is thought to depend on connections with the cerebral cortex. Anatomical investigations in animals and post-mortem humans have established that cerebro-cerebellar connections are contralateral to each other and include the cerebello-thalamo-cortical (CTC) and cortico-ponto-cerebellar (CPC) pathways. CTC and CPC characterization in humans in vivo is still challenging. Here advanced tractography was combined with quantitative indices to compare CPC to CTC pathways in healthy subjects. Differently to previous studies, our findings reveal that cerebellar cognitive areas are reached by the largest proportion of the reconstructed CPC, supporting the hypothesis that a CTC-CPC loop provides a substrate for cerebro-cerebellar communication during cognitive processing. Amongst the cerebral areas identified using in vivo tractography, in addition to the cerebral motor cortex, major portions of CPC streamlines leave the prefrontal and temporal cortices. These findings are useful since provide MRI-based indications of possible subtending connectivity and, if confirmed, they are going to be a milestone for instructing computational models of brain function. These results, together with further multi-modal investigations, are warranted to provide important cues on how the cerebro-cerebellar loops operate and on how pathologies involving cerebro-cerebellar connectivity are generated.

In addition to motor functions, it has become clear that in humans the cerebellum plays a significant role in cognition too, through connections with associative areas in the cerebral cortex. Classical anatomy indicates that neo-cerebellar regions are connected with the contralateral cerebral cortex through the dentate nucleus, superior cerebellar peduncle, red nucleus and ventrolateral anterior nucleus of the thalamus. The anatomical existence of these connections has been demonstrated using virus retrograde transport techniques in monkeys and rats ex vivo. In this study, using advanced diffusion MRI tractography we show that it is possible to calculate streamlines to reconstruct the pathway connecting the cerebellar cortex with contralateral cerebral cortex in humans in vivo. Corresponding areas of the cerebellar and cerebral cortex encompassed similar proportion (about 80 %) of the tract, suggesting that the majority of streamlines passing through the superior cerebellar peduncle connect the cerebellar hemispheres through the ventrolateral thalamus with contralateral associative areas. This result demonstrates that this kind of tractography is a useful tool to map connections between the cerebellum and the cerebral cortex and moreover could be used to support specific theories about the abnormal communication along these pathways in cognitive dysfunctions in pathologies ranging from dyslexia to autism.Electronic supplementary materialThe online version of this article (doi:10.1007/s00429-014-0861-2) contains supplementary material, which is available to authorized users.

The projectional nature of radiogram limits its amount of information about the instrumented spine. MRI and CT imaging can be more helpful, using cross-sectional view. However, the presence of metal-related artifacts at both conventional CT and MRI imaging can obscure relevant anatomy and disease. We reviewed the literature about overcoming artifacts from metallic orthopaedic implants at high-field strength MRI imaging and multi-detector CT. The evolution of multichannel CT has made available new techniques that can help minimizing the severe beam-hardening artifacts. The presence of artifacts at CT from metal hardware is related to image reconstruction algorithm (filter), tube current (in mA), X-ray kilovolt peak, pitch, hardware composition, geometry (shape), and location. MRI imaging has been used safely in patients with orthopaedic metallic implants because most of these implants do not have ferromagnetic properties and have been fixed into position. However, on MRI imaging metallic implants may produce geometric distortion, the so-called susceptibility artifact. In conclusion, although 140 kV and high milliamperage second exposures are recommended for imaging patients with hardware, caution should always be exercised, particularly in children, young adults, and patients undergoing multiple examinations. MRI artifacts can be minimized by positioning optimally and correctly the examined anatomy part with metallic implants in the magnet and by choosing fast spin-echo sequences, and in some cases also STIR sequences, with an anterior to posterior frequencyencoding direction and the smallest voxel size.

Dynamic susceptibility contrast magnetic resonance imaging (DSC-MRI) is used to track the first pass of an exogenous, paramagnetic, nondiffusible contrast agent through brain tissue, and has emerged as a powerful tool in the characterization of brain tumor hemodynamics. DSC-MRI parameters can be helpful in many aspects, including tumor grading, prediction of treatment response, likelihood of malignant transformation, discrimination between tumor recurrence and radiation necrosis, and differentiation between true early progression and pseudoprogression. This review aims to provide a conceptual overview of the underlying principles of DSC-MRI of the brain for clinical neuroradiologists, scientists, or students wishing to improve their understanding of the technical aspects, pitfalls, and controversies of DSC perfusion MRI of the brain. Future consensus on image acquisition parameters and postprocessing of DSC-MRI will most likely allow this technique to be evaluated and used in high-quality multicenter studies and ultimately help guide clinical care.

In resting state fMRI (rs-fMRI), only functional connectivity (FC) reductions in the default mode network (DMN) are normally reported as a biomarker for Alzheimer's disease (AD). In this investigation we have developed a comprehensive strategy to characterize the FC changes occurring in multiple networks and applied it in a pilot study of subjects with AD and Mild Cognitive Impairment (MCI), compared to healthy controls (HC). Resting state networks (RSNs) were studied in 14 AD (70 ± 6 years), 12 MCI (74 ± 6 years), and 16 HC (69 ± 5 years). RSN alterations were present in almost all the 15 recognized RSNs; overall, 474 voxels presented a reduced FC in MCI and 1244 in AD while 1627 voxels showed an increased FC in MCI and 1711 in AD. The RSNs were then ranked according to the magnitude and extension of FC changes (gFC), putting in evidence 6 RSNs with prominent changes: DMN, frontal cortical network (FCN), lateral visual network (LVN), basal ganglia network (BGN), cerebellar network (CBLN), and the anterior insula network (AIN). Nodes, or hubs, showing alterations common to more than one RSN were mostly localized within the prefrontal cortex and the mesial-temporal cortex. The cerebellum showed a unique behavior where voxels of decreased gFC were only found in AD while a significant gFC increase was only found in MCI. The gFC alterations showed strong correlations (p < 0.001) with psychological scores, in particular Mini-Mental State Examination (MMSE) and attention/memory tasks. In conclusion, this analysis revealed that the DMN was affected by remarkable FC increases, that FC alterations extended over several RSNs, that derangement of functional relationships between multiple areas occurred already in the early stages of dementia. These results warrant future work to verify whether these represent compensatory mechanisms that exploit a pre-existing neural reserve through plasticity, which evolve in a state of lack of connectivity between different networks with the worsening of the pathology.

Brain function has been investigated via the blood oxygenation level dependent (BOLD) effect using magnetic resonance imaging (MRI) for the past decades. Advances in sodium imaging offer the unique chance to access signal changes directly linked to sodium ions (23Na) flux across the cell membrane, which generates action potentials, hence signal transmission in the brain. During this process 23Na transiently accumulates in the intracellular space. Here we show that quantitative functional sodium imaging (fNaI) at 3T is potentially sensitive to 23Na concentration changes during finger tapping, which can be quantified in gray and white matter regions key to motor function. For the first time, we measured a 23Na concentration change of 0.54 mmol/l in the ipsilateral cerebellum, 0.46 mmol/l in the contralateral primary motor cortex (M1), 0.27 mmol/l in the corpus callosum and -11 mmol/l in the ipsilateral M1, suggesting that fNaI is sensitive to distributed functional alterations. Open issues persist on the role of the glymphatic system in maintaining 23Na homeostasis, the role of excitation and inhibition as well as volume distributions during neuronal activity. Haemodynamic and physiological signal recordings coupled to realistic models of tissue function will be critical to understand the mechanisms of such changes and contribute to meeting the overarching challenge of measuring neuronal activity in vivo.

Among dementia-like diseases, Alzheimer disease (AD) and vascular dementia (VD) are two of the most frequent. AD and VD may share multiple neurological symptoms that may lead to controversial diagnoses when using conventional clinical and MRI criteria. Therefore, other approaches are needed to overcome this issue. Machine learning (ML) combined with magnetic resonance imaging (MRI) has been shown to improve the diagnostic accuracy of several neurodegenerative diseases, including dementia. To this end, in this study, we investigated, first, whether different kinds of ML algorithms, combined with advanced MRI features, could be supportive in classifying VD from AD and, second, whether the developed approach might help in predicting the prevalent disease in subjects with an unclear profile of AD or VD. Three ML categories of algorithms were tested: artificial neural network (ANN), support vector machine (SVM), and adaptive neuro-fuzzy inference system (ANFIS). Multiple regional metrics from resting-state fMRI (rs-fMRI) and diffusion tensor imaging (DTI) of 60 subjects (33 AD, 27 VD) were used as input features to train the algorithms and find the best feature pattern to classify VD from AD. We then used the identified VD-AD discriminant feature pattern as input for the most performant ML algorithm to predict the disease prevalence in 15 dementia patients with a "mixed VD-AD dementia" (MXD) clinical profile using their baseline MRI data. ML predictions were compared with the diagnosis evidence from a 3-year clinical follow-up. ANFIS emerged as the most efficient algorithm in discriminating AD from VD, reaching a classification accuracy greater than 84% using a small feature pattern. Moreover, ANFIS showed improved classification accuracy when trained with a multimodal input feature data set (e.g., DTI + rs-fMRI metrics) rather than a unimodal feature data set. Castellazzi et al. Dementia Classification Using Machine Learning When applying the best discriminant pattern to the MXD group, ANFIS achieved a correct prediction rate of 77.33%. Overall, results showed that our approach has a high discriminant power to classify AD and VD profiles. Moreover, the same approach also showed potential in predicting earlier the prevalent underlying disease in dementia patients whose clinical profile is uncertain between AD and VD, therefore suggesting its usefulness in supporting physicians' diagnostic evaluations.

Alzheimer disease (AD) and vascular dementia (VaD) together represent the majority of dementia cases. Since their neuropsychological profiles often overlap and white matter lesions are observed in elderly subjects including AD, differentiating between VaD and AD can be difficult. Characterization of these different forms of dementia would benefit by identification of quantitative imaging biomarkers specifically sensitive to AD or VaD. Parameters of microstructural abnormalities derived from diffusion tensor imaging (DTI) have been reported to be helpful in differentiating between dementias, but only few studies have used them to compare AD and VaD with a voxelwise approach. Therefore, in this study a whole brain statistical analysis was performed on DTI data of 93 subjects (31 AD, 27 VaD, and 35 healthy controls—HC) to identify specific white matter patterns of alteration in patients affected by VaD and AD with respect to HC. Parahippocampal tracts were found to be mainly affected in AD, while VaD showed more spread white matter damages associated with thalamic radiations involvement. The genu of the corpus callosum was predominantly affected in VaD, while the splenium was predominantly affected in AD revealing the existence of specific patterns of alteration useful in distinguishing between VaD and AD. Therefore, DTI parameters of these regions could be informative to understand the pathogenesis and support the etiological diagnosis of dementia. Further studies on larger cohorts of subjects, characterized for brain amyloidosis, will allow to confirm and to integrate the present findings and, furthermore, to elucidate the mechanisms of mixed dementia. These steps will be essential to translate these advances to clinical practice.