Algorithms for computer-aided diagnosis of dementia based on structural MRI have demonstrated high performance in the literature, but are difficult to compare as different data sets and methodology were used for evaluation. In addition, it is unclear how the algorithms would perform on previously unseen data, and thus, how they would perform in clinical practice when there is no real opportunity to adapt the algorithm to the data at hand. To address these comparability, generalizability and clinical applicability issues, we organized a grand challenge that aimed to objectively compare algorithms based on a clinically representative multi-center data set. Using clinical practice as starting point, the goal was to reproduce the clinical diagnosis. Therefore, we evaluated algorithms for multi-class classification of three diagnostic groups: patients with probable Alzheimer’s disease, patients with mild cognitive impairment and healthy controls. The diagnosis based on clinical criteria was used as reference standard, as it was the best available reference despite its known limitations. For evaluation, a previously unseen test set was used consisting of 354 T1-weighted MRI scans with the diagnoses blinded. Fifteen research teams participated with in total 29 algorithms. The algorithms were trained on a small training set (n=30) and optionally on data from other sources (e.g., the Alzheimer’s Disease Neuroimaging Initiative, the Australian Imaging Biomarkers and Lifestyle flagship study of aging). The best performing algorithm yielded an accuracy of 63.0% and an area under the receiver-operating-characteristic curve (AUC) of 78.8%. In general, the best performances were achieved using feature extraction based on voxel-based morphometry or a combination of features that included volume, cortical thickness, shape and intensity. The challenge is open for new submissions via the web-based framework: http://caddementia.grand-challenge.org.

Introduction: Imageability is a psycholinguistic variable that indicates how well a word gives rise to a mental image or sensory experience. Imageability ratings are used extensively in psycholinguistic, neuropsychological and aphasiological studies. However, little formal knowledge exists on whether and how these ratings are associated between and within languages. Methods and results:Fifteen imageability databases were cross-correlated using non-parametric statistics. Some of these corresponded to unpublished data collected within a European research network -the Collaboration of Aphasia Trialists (COST IS1208). All but four correlations were significant. The average strength of the correlations (rho =.68) and the variance explained (R 2 =46%)were moderate. This implies that factors other than imageability may explain 54% of the results. Conclusion: Imageability ratings often correlate across languages. Different possibly interactingfactors may explain the moderate strength and variance in the correlations: (1) linguistic and cultural factors; (2) intrinsic differences between databases; (3) range effects; (4) small numbers of words in each database, equivalent words, and participants; and (5) mean age of participants. The results suggest that imageability ratings may be used cross-linguistically. However, further understanding of the factors explaining the variance in the correlations is needed, before research and practice recommendations can be made.

BACKGROUND AND PURPOSE:Determining language dominance with fMRI is challenging in patients with brain tumor, particularly in cases of suspected atypical language representation. Supratentorial activation patterns must be interpreted with great care when the tumor is in or near the presumed language areas, where tumor tissue or mass effect can lead to false-negative fMRI results. In this study, we assessed cerebrocerebellar language fMRI lateralization in healthy participants and in patients with brain tumors with a focus on atypical language representation.