Recurrence quantification analysis, variability analysis, and fractal dimension estimation in 99MTc-HDP nuclear scintigraphy of maxillary bones in subjects with osteoporosis

We develop a non linear methodology to analyze nuclear medicine images. It holds on Recurrence Quantification Analysis (RQA), on Analysis of Variability by variogram, and on estimation of Fractal Dimension. It is applied to five subjects with osteoporosis in comparison with five control subjects. Bone nuclear images are obtained after administration of 99mTc-HDP. Regions of interest (ROI) are selected in maxillary bones of oral cavity. Some basic non linear indices are obtained as result of the methodology and they enable quantitative estimations at the micro-structural and micro-architectural level of bone matrix investigated. The indices result very satisfactory in discriminating controls from subjects with osteoporosis. Their appear of interest also in the case of dentist intervention often engaged to evaluate oral signs and, in particular, to utilize mandibular or maxillary bones indices in relation to possible loss of bone mineral density and/or to microarchitectural deterioration of bone tissue.