作者: G. De Vico , M. Cataldi , P. Maiolino , S. Beltraminelli , G.A. Losa
关键词: Magnification 、 Fractal pattern 、 Mathematics 、 Fractal 、 Pattern recognition 、 Artificial intelligence 、 Box counting 、 Trichoblastoma 、 Fractal dimension 、 Fractal analysis
摘要: Trichoblastoma, is a “benign tumour derived from or reduplicating the primitive hair germ of embrionic follicular development”. Among different subtypes canine trichoblastoma, Ribbon type Trichoblastoma subjectively display very complex structure, sometimes suggestive self-similar design, which well known characteristic fractal structures. In this study, we performed analysis twelve (12) trichoblastoma in order to test use approach characterise and describe architecture epithelial growth peculiar spontaneous tumour.For each case, was basically by FANAL ++ software, determines slope possible region bi-asymptotic curve. The FANAL++ results were also compared with dimension (FD) calculated on same images box counting method further commercially available software (Benoit 1.3) doesn’t extrapolate windows tumour. cases examined our data confirm that subjective self-similarity observed component reflect true pattern. Furthermore, values FD Benoit 1.3 are general comparable statistical point view, although numerically different, depending closely settings. Finally, demonstrated difference when captured at magnification. particular increased magnification decreased. authors suggest these findings should be considered for standardised based classification trichoblastoma.