Shape-based disease grading via functional maps and graph convolutional networks with application to Alzheimer’s disease

• Published in: BMC medical imaging Vol. 24; no. 1; pp. 342 - 10
• Main Authors: Mayer, Julius, Baum, Daniel, Ambellan, Felix, von Tycowicz, Christoph
• Format: Journal Article
• Language: English
• Published: London BioMed Central 18.12.2024; Springer Nature B.V; BMC
• Subjects: Aged; Alzheimer Disease - diagnostic imaging; Alzheimer's disease; Artificial neural networks; Classification; Computer-aided diagnosis; Concept learning; Deep Learning; Disease control; Geometric deep learning; Geometry; Humans; Image Interpretation, Computer-Assisted - methods; Imaging; Machine learning; Magnetic Resonance Imaging - methods; Mathematical functions; Medical imaging; Medicine; Medicine & Public Health; Neural networks; Neural Networks, Computer; Neurodegenerative diseases; Performance evaluation; Radiology; Shape analysis; Geometric deep learning; Shape analysis; Alzheimer’s disease; Computer-aided diagnosis
• ISSN: 1471-2342; 1471-2342
• DOI: 10.1186/s12880-024-01513-z

Shape analysis provides methods for understanding anatomical structures extracted from medical images. However, the underlying notions of shape spaces that are frequently employed come with strict assumptions prohibiting the analysis of incomplete and/or topologically varying shapes. This work aims to alleviate these limitations by adapting the concept of functional maps . Further, we present a graph-based learning approach for morphometric classification of disease states that uses novel shape descriptors based on this concept. We demonstrate the performance of the derived classifier on the open-access ADNI database differentiating normal controls and subjects with Alzheimer’s disease. Notably, the experiments show that our approach can improve over state-of-the-art from geometric deep learning.