Abstract #0579

Roy R, Srivastava M, Pandey A, Ramakant P, Kumar S, Srivastava A

PROTON MR quantitative analysis in breast tumors, fibroadenomas (n=10) and ductal carcinomas (n=13) along with their respective controls (n=23) were carried out on twelve metabolites viz. formate, myo-inositol, glycine, a-GPC, choline, Cr/PCr, glutamine, succinate, glutamate, acetate, alanine and lactate. Multivariate discriminant function analysis (DFA) was performed to find out important descriptors for the discrimination of fibroadenomas and ductal carcinomas (CA), along with their control tissues. The best discrimination between fibroadenomas and ductal carcinomas was obtained with the combined effect of Glycine and Choline as marker metabolites (p< 0.01). The proposed model could classify 95.65% cases correctly between the fibroadenoma and ductal carcinoma, with the sensitivity of 92.3% and the specificity of 100%.