Stéphane Lehéricy

Automatic classification of patients with Alzheimer's disease from structural MRI: A comparison of ten methods using the ADNI database

Recently, several high dimensional classification methods have been proposed to automatically discriminate between patients with Alzheimers disease (AD) or mild cognitive impairment (MCI) and elderly controls (CN) based on T1-weighted MRI. However, these methods were assessed on different populations, making it difficult to compare their performance. In this paper, we evaluated the performance of ten approaches (five voxel-based methods, three methods based on cortical thickness and two methods based on the hippocampus) using 509 subjects from the ADNI database. Three classification experiments were performed: CN vs AD, CN vs MCIc (MCI who had converted to AD within 18 months, MCI converters - MCIc) and MCIc vs MCInc (MCI who had not converted to AD within 18 months, MCI non-converters - MCInc). Data from 81 CN, 67 MCInc, 39 MCIc and 69 AD were used for training and hyperparameters optimization. The remaining independent samples of 81 CN, 67 MCInc, 37 MCIc and 68 AD were used to obtain an unbiased estimate of the performance of the methods. For AD vs CN, whole-brain methods (voxel-based or cortical thickness-based) achieved high accuracies (up to 81% sensitivity and 95% specificity). For the detection of prodromal AD (CN vs MCIc), the sensitivity was substantially lower. For the prediction of conversion, no classifier obtained significantly better results than chance. We also compared the results obtained using the DARTEL registration to that using SPM5 unified segmentation. DARTEL significantly improved six out of 20 classification experiments and led to lower results in only two cases. Overall, the use of feature selection did not improve the performance but substantially increased the computation times.

Goal-directed and habitual control in the basal ganglia: implications for Parkinson's disease

Progressive loss of the ascending dopaminergic projection in the basal ganglia is a fundamental pathological feature of Parkinsons disease. Studies in animals and humans have identified spatially segregated functional territories in the basal ganglia for the control of goal-directed and habitual actions. In patients with Parkinsons disease the loss of dopamine is predominantly in the posterior putamen, a region of the basal ganglia associated with the control of habitual behaviour. These patients may therefore be forced into a progressive reliance on the goal-directed mode of action control that is mediated by comparatively preserved processing in the rostromedial striatum. Thus, many of their behavioural difficulties may reflect a loss of normal automatic control owing to distorting output signals from habitual control circuits, which impede the expression of goal-directed action.

Anatomical variability in the cortical representation of first and second language

FUNCTIONAL magnetic resonance imaging was used to assess inter-subject variability in the cortical representation of language comprehension processes. Moderately fluent French-English bilinguals were scanned while they listened to stories in their first language (L1 = French) or in a second language (L2 = English) acquired at school after the age of seven. In all subjects, listening to L1 always activated a similar set of areas in the left temporal lobe, clustered along the left superior temporal sulcus. Listening to L2, however, activated a highly variable network of left and right temporal and frontal areas, sometimes restricted only to right-hemispheric regions. These results support the hypothesis that first language acquisition relies on a dedicated left-hemispheric cerebral network, while late second language acquisition is not necessarily associated with a reproducible biological substrate. The postulated contribution of the right hemisphere to L2 comprehension1 is found to hold only on average, individual subjects varying from complete right lateralization to standard left lateralization for L2.

Diffusion tensor fiber tracking shows distinct corticostriatal circuits in humans

A landmark of corticostriatal connectivity in nonhuman primates is that cortical connections are organized into a set of discrete circuits. Each circuit is assumed to perform distinct behavioral functions. In animals, most connectivity studies are performed using invasive tracing methods, which are nonapplicable in humans. To test the proposal that corticostriatal connections are organized as multiple circuits in humans, we used diffusion tensor imaging axonal tracking, a new magnetic resonance technique that allows demonstration of fiber tracts in a noninvasive manner. Diffusion tensor imaging–based fiber tracking showed that the posterior (sensorimotor), anterior (associative), and ventral (limbic) compartments of the human striatum have specific connections with the cortex, and particularly the frontal lobes. These results provide the first direct demonstration of distinct corticostriatal connections in humans.

Contributions of the basal ganglia and functionally related brain structures to motor learning

This review discusses the cerebral plasticity, and the role of the cortico-striatal system in particular, observed as one is learning or planning to execute a newly learned motor behavior up to when the skill is consolidated or has become highly automatized. A special emphasis is given to imaging work describing the neural substrate mediating motor sequence learning and motor adaptation paradigms. These results are then put into a plausible neurobiological model of motor skill learning, which proposes an integrated view of the brain plasticity mediating this form of memory at different stages of the acquisition process.

Functional MR evaluation of temporal and frontal language dominance compared with the Wada test

Objective: To evaluate the reliability of temporal and frontal functional MRI (fMRI) activation for the assessment of language dominance, as compared with the Wada test. Patients and Methods: Ten patients with temporal lobe epilepsy were studied using blood oxygen level dependent fMRI and echoplanar imaging (1.5-T). Three tasks were used: semantic verbal fluency, covert sentence repetition, and story listening. Data were analyzed using pixel by pixel autocorrelation and cross-correlation. fMRI laterality indices were defined for several regions of interest as the ratio (L − R)/(L + R), L being the number of activated voxels in the left hemisphere and R in the right hemisphere. Wada laterality indices were defined as the difference in the percentages of errors in language tests between left and right carotid injections. Results: Semantic verbal fluency: The asymmetry of frontal activation was correlated with Wada laterality indices. The strongest correlation was observed in the precentral/middle frontal gyrus/inferior frontal sulcus area. Story listening: The asymmetry of frontal, but not temporal, activation was correlated with Wada laterality indices. Covert sentence repetition: No correlation was observed. Conclusions: There was a good congruence between hemispheric dominance for language as assessed with the Wada test and fMRI laterality indices in the frontal but not in the temporal lobes. The story listening and the covert sentence repetition tasks increased the sensitivity of detection of posterior language sites that may be useful for brain lesion surgery.

The neural system that bridges reward and cognition in humans: An fMRI study

We test the hypothesis that motivational and cognitive processes are linked by a specific neural system to reach maximal efficiency. We studied six normal subjects performing a working memory paradigm (n-back tasks) associated with different levels of monetary reward during an fMRI session. The study showed specific brain activation in relation with changes in both the cognitive loading and the reward associated with task performance. First, the working memory tasks activated a network including the dorsolateral prefrontal cortex [Brodmann area (BA) 9/46] and, in addition, in the lateral frontopolar areas (BA 10), but only in the more demanding condition (3-back task). This result suggests that lateral prefrontal areas are organized in a caudo-rostral continuum in relation with the increase in executive requirement. Second, reward induces an increased activation in the areas already activated by working memory processing and in a supplementary region, the medial frontal pole (BA 10), regardless of the level of cognitive processing. It is postulated that the latter region plays a specific role in monitoring the reward value of ongoing cognitive processes. Third, we detected areas where the signal decreases (ventral-BA 11/47 and subgenual prefrontal cortices) in relation with both the increase of cognitive demand and the reward. The deactivation may represent an emotional gating aimed at inhibiting adverse emotional signals to maximize the level of performance. Taken together, these results suggest a balance between increasing activity in cortical cognitive areas and decreasing activity in the limbic and paralimbic structures during ongoing higher cognitive processing.

Partial correlation for functional brain interactivity investigation in functional MRI.

Examination of functional interactions through effective connectivity requires the determination of three distinct levels of information: (1) the regions involved in the process and forming the spatial support of the network, (2) the presence or absence of interactions between each pair of regions, and (3) the directionality of the existing interactions. While many methods exist to select regions (Step 1), very little is available to complete Step 2. The two main methods developed so far, structural equation modeling (SEM) and dynamical causal modeling (DCM), usually require precise prior information to be used, while such information is sometimes lacking. Assuming that Step 1 was successfully completed, we here propose a data-driven method to deal with Step 2 and extract functional interactions from fMRI datasets through partial correlations. Partial correlation is more closely related to effective connectivity than marginal correlation and provides a convenient graphical representation for functional interactions. As an instance of brain interactivity investigation, we consider how simple hand movements are processed by the bihemispheric cortical motor network. In the proposed framework, Bayesian analysis makes it possible to estimate and test the partial statistical dependencies between regions without any prior model on the underlying functional interactions. We demonstrate the interest of this approach on real data.

Cholinergic mesencephalic neurons are involved in gait and postural disorders in Parkinson disease.

Gait disorders and postural instability, which are commonly observed in elderly patients with Parkinson disease (PD), respond poorly to dopaminergic agents used to treat other parkinsonian symptoms. The brain structures underlying gait disorders and falls in PD and aging remain to be characterized. Using functional MRI in healthy human subjects, we have shown here that activity of the mesencephalic locomotor region (MLR), which is composed of the pedunculopontine nucleus (PPN) and the adjacent cuneiform nucleus, was modulated by the speed of imagined gait, with faster imagined gait activating a discrete cluster within the MLR. Furthermore, the presence of gait disorders in patients with PD and in aged monkeys rendered parkinsonian by MPTP intoxication correlated with loss of PPN cholinergic neurons. Bilateral lesioning of the cholinergic part of the PPN induced gait and postural deficits in nondopaminergic lesioned monkeys. Our data therefore reveal that the cholinergic neurons of the PPN play a central role in controlling gait and posture and represent a possible target for pharmacological treatment of gait disorders in PD.

Multidimensional classification of hippocampal shape features discriminates Alzheimer's disease and mild cognitive impairment from normal aging

We describe a new method to automatically discriminate between patients with Alzheimers disease (AD) or mild cognitive impairment (MCI) and elderly controls, based on multidimensional classification of hippocampal shape features. This approach uses spherical harmonics (SPHARM) coefficients to model the shape of the hippocampi, which are segmented from magnetic resonance images (MRI) using a fully automatic method that we previously developed. SPHARM coefficients are used as features in a classification procedure based on support vector machines (SVM). The most relevant features for classification are selected using a bagging strategy. We evaluate the accuracy of our method in a group of 23 patients with AD (10 males, 13 females, age+/-standard-deviation (SD)=73+/-6 years, mini-mental score (MMS)=24.4+/-2.8), 23 patients with amnestic MCI (10 males, 13 females, age+/-SD=74+/-8 years, MMS=27.3+/-1.4) and 25 elderly healthy controls (13 males, 12 females, age+/-SD=64+/-8 years), using leave-one-out cross-validation. For AD vs controls, we obtain a correct classification rate of 94%, a sensitivity of 96%, and a specificity of 92%. For MCI vs controls, we obtain a classification rate of 83%, a sensitivity of 83%, and a specificity of 84%. This accuracy is superior to that of hippocampal volumetry and is comparable to recently published SVM-based whole-brain classification methods, which relied on a different strategy. This new method may become a useful tool to assist in the diagnosis of Alzheimers disease.