Differentiation of focal hepatic lesions in MR imaging with the use of combined quantitative and qualitative analysis

Background: To evaluate the efficacy of the protocol of combined quantitative-qualitative analysis for the differentiation of focal hepatic lesions. Material/Methods: The study group included 168 patients with 292 hepatic lesions confirmed by histology (n =138) or follow-up (n =154). Lesions were divided into: benign lesions treated conservatively (group A, 120 lesions), malignant tumors and benign lesions treated surgically (group B, 172 lesions). MR imaging (1.5-T) consisted of sequences: T2 double-echo TSE, T2 STIR, T1 GRE and of dynamic study. During the first part of differentiation process, quantitative analysis, based on lesions T2 relaxation times (derived from T2 double-echo TSE sequence), was performed in order to discriminate non-solid lesions (hemangiomas, cysts, abscesses; n = 88) from solid tumors (n = 204). Subsequently, all tumors defined as solid underwent qualitative evaluation based on visual assessment of lesions signal intensities in all sequences and patterns of their contrast enhancement. The aim of this part of analysis was to discriminate benign lesions (FNH and focal fatty infiltration) from other solid tumors. The remaining tumors were characterized as group B lesions. Results: Statistically significant difference between mean T2 relaxation time of solid tumors (84.1 ms) and non-solid lesions (250.5 ms) was noted, allowing diagnosis of solid tumors with sensitivity of 96% and specificity of 93% (at the threshold of 116 ms). Overall 202 lesions were defined as solid (196 true positive, 8 false negative, 6 false positive results). Qualitative analysis of these lesions was performed allowing correct characterization of all 7 focal fatty infiltrations and 21 of 24 FNH. Six lesions were falsely diagnosed as FNH. Remaining 168 lesions were defined as group B lesions. Both parts of differentiation protocol yielded sensitivity and specificity of 92%, allowing correct characterization of 158 of 172 group B lesions. Fourteen false negative and 10 false positive results (3 FNH, 1 focal inflammation, 6 hemangiomas) were obtained. Conclusions: Combined protocol of quantitative and qualitative analysis enabled discrimination of group B lesions (malignant tumors and benign lesions treated surgically) with high sensitivity and specificity of 92%.