Daniele De Falco Alfano
University of Ferrara, Italy
Title: Texture Analysis of clear cell renal cell carcinoma
Biography
Biography: Daniele De Falco Alfano
Abstract
Purpose: The main goal of this retrospective study in blind, was to evaluate the accuracy of Mean value of Positive Pixels (MPP) texture parameter in tomodensitometric evaluation of renal cell carcinoma. Materials & Methods: IRB approval was obtained for the study. All patients who underwent a surgical nephrectomy (radical or partial) between January and December of 2013 has been identified. The study sample consisted of 33 men and 16 women between the age of 37 and 91 years. The TCTA study was conducted using the commercial software Image J, National Institutes of Health, USA. Result: A statistically significant correlation between MPP values (rs=0.66; p<0.001) with Histologic Grading (Gold Standard) was documented. The average values of MPP were significantly lower in the non-tumor lesions (MPP 92.63±3.81) compared to well-differentiated lesions G1 (MPP=112.35±4.03), moderately differentiated lesions G2 (MPP = 142.41±13.83; ) and G3 (PP = 147.21±10.74) and to poorly differentiated lesions G4 (MPP =177.89±6.57). The intra-observer and interobserver accordande was excellent (respectively a=0.94 and 0.96). Conclusion: The CT Texture Analysis is a valid aid for the characterization of tumor lesions allowing a clear distinction of well differentiated carcinomas (G1), moderately differentiated (G2 and G3) and poorly differentiated (G4). The evaluation of MPP texture parameter represent an accurate and reproducible approach for tomodensitometric characterization of renal cell carcinomas.
Speaker Presentations