作者: P. KATSALOULIS , D. A. VERGANELAKIS , A. PROVATA
DOI: 10.1142/S0218348X09004284
关键词: Mathematics 、 Pattern recognition 、 Artificial intelligence 、 Fractal dimension 、 Topology 、 Cluster analysis 、 Standard deviation 、 Box counting 、 Human brain 、 Tractography 、 Lacunarity
摘要: Tractography images produced by Magnetic Resonance Imaging scans have been used to calculate the topology of neuron tracts in human brain. This technique gives neuroanatomical details, limited system resolution properties. In observed scales demonstrated statistical self-similar structure axons and its fractal dimensions were estimated using classic Box Counting technique. To assess degree clustering neural network, lacunarity was calculated Gliding method. The two-dimensional tractography taken from four subjects various angles different parts results that average dimension is approximately Df = 1.60 with standard deviation 0.11 for healthy human-brain tissues, it presents self-similarity features similar many other biological root-like structures.