Rancang Bangun Agregator Toko Aplikasi Mobile Berbasis Web Untuk Menyediakan Informasi Kompatibilitas Aplikasi Multi Platform

Abstract

Objective.To detect the structural brain signatures of patients with a late age-of-onset major depressive disorder (lo-MDD) relative to controls (training-dataset), and to test the ability of such measures to predict the presence of depression in Alzheimer's disease (AD) and behavioral variant of frontotemporal dementia (bvFTD) (testing-dataset). Background.Whether depressive symptoms are intrinsic to neurodegeneration is still poorly understood. Lo-MDD may represent a model to understand the nature of depressive symptoms in AD and bvFTD. Methods.T1-weighted and diffusion tensor (DT)-MRI were obtained from 15 lo-MDD and 28 controls, and from an age-matched dataset of 61 AD and 27 bvFTD patients. Participants underwent a clinical assessment to detect the presence of depression. Cortical thickness (CT) and white matter (WM) metrics were obtained from all subjects, and compared among training-dataset groups to identify signatures of depression. MRI measures found to be significantly different between training-dataset groups were used for an individual classification of depressed/non-depressed subjects of the testing-dataset using a ROC-curve analysis. Results.Compared with controls, lo-MDD showed cortical thinning of the bilateral superior and middle-temporal cortices and rostral/middle-frontal and orbitofrontal cortices, and altered WM metrics of the corpus callosum and right parahippocampal bundle. Depressive symptoms were detected in 56[percnt] of AD and 52[percnt] of bvFTD patients. In AD, the best model to detect depression was a model combining the CT measures ("GM-model": AUC=0.73; sensitivity=0.71; specificity=0.74), while in bvFTD the best prediction was achieved by combining CT and DT-MRI metrics ("GM+WM-model": AUC=0.92; sensitivity=0.83; specificity=0.91). Conclusions.Lo-MDD is a good model to understand the nature of depression in AD and bvFTD, which is likely to be inherently associated with neurodegeneration in both diseases. The discrimination accuracies obtained suggest that models combining CT measures in AD and both CT and WM metrics in bvFTD are potentially relevant for depression detection in these disorders.