FRACTAL SURFACE-ANALYSIS OF PULMONARY NOD ULES BASED ON HIGH-RESOLUTION COMPUTED-TOMOGRAPHY

Purpose: The visual assessment of shape and surface structures of pulm onary nodules (PN) (smooth edges vs. spiculae) serves as a qualitative , subjective criterion in differential diagnosis. These properties mus t be quantified by computer-assisted evaluation, an adequate mathemati cal model, and new quantitative shape parameters. Methods: 12 patients were investigated by high resolution CT. Based on 3D reconstructions with increasing thresholds, the TN surface S, volume V, fractal dimens ion (FD=ln(S)/ln((3) root V)), and fractal index (Fl=root S/(3) root V ) were calculated. Results: The relations between the reconstruction t hreshold and the calculated tumour surface, respective volume, are of a fractal nature: S=c(1) x thres(1/d), V=c(2) x thres(1/D) (c(1), c(2) , d, D: real constants). Whereas the absolute values of surface and vo lume strongly depend on the chosen threshold (mean volume differences of 249%), the derived parameter of the fractal dimension for a specifi c PN is nearly constant for all thresholds under review (r=0.998, SD=0 .05). Conclusions: FD and FI are new diagnostic features for the asses sment of PN surface structure and morphology. Thus, the assessment of pulmonary nodules can be supported by new quantitative parameters repr esenting surface irregularity due to invasive tumour growth into adjac ent tissue.