Classificação da doença de Alzheimer baseada em graph kernels construídos a partir de atributos de textura 3D de imagens de ressonância magnética

• Autor(a) principal: Mendonça, Lucas José Cruz de
• Data de Publicação: 2022
• Tipo de documento: Dissertação
• Idioma: por
• Título da fonte: Repositório Institucional da UFSCAR
• Texto Completo: https://repositorio.ufscar.br/handle/ufscar/15789
• topic: Doença de Alzheimer; Classificação de grafos; Extração de textura; Imagem de ressonância magnética; Alzheimer's disease; Graphs-based classification; Texture extraction; Magnetic resonance image; Graph kernel; CIENCIAS EXATAS E DA TERRA::CIENCIA DA COMPUTACAO

Alzheimer's disease (AD) is neurodegenerative and characterized by cognitive and behavioral impairment. Mild cognitive impairment (MCI) is a relatively broad clinical condition involving a slight memory deficit, which in many cases represents a transitional state between a cognitively normal (CN) condition and AD. Many published studies restrict their analyses to searching for structural changes caused by the disease in a few particular regions of the brain.. Currently, the studies are looking for new AD biomarkers using multiple brain regions and focusing on subtle texture changes in the image. Therefore, this study proposes a new technique for MR image classification in AD diagnosis using graph kernels constructed from texture features extracted from 3D Structural magnetic resonance (sMR) images. In our proposed method we first segment the brain images into multiple regions with the FreeSurfer. Then, we extract 22 texture features and defined the graph node attributes as the probability distributions of the extracted features. Next, for each texture feature we build a graph and define its edge weights as the distances between pairs of node attributes using three metrics. After that, we use a threshold-based approach for graph edges removal. Finally, we perform graph classification using Support Vector Machines (SVMs) with two graph-kernels. Results of our method have shown better performances for the CN×AD (AUC=0.92) and CN×MCI (AUC=0.81) classifications, and worse for MCI×AD (AUC=0.78). This trend is consistent with other published results and makes sense if we consider the concept of Alzheimer's disease continuum from pathophysiological, biomarker and clinical perspectives. Besides allowing the use of different texture attributes for the diagnosis of AD, our method uses the graph-kernel approach to represent texture features from different regions of the brain image, which considerably facilitates the image classification task via SVMs. Our results were promising when compared to the state-of-the-art in graph-based AD classification.