Abstract #1537

Interobserver Agreement of Semi-Automated, Volumetric Analysis of Multiparametric MRI in Patients with Liver Cancer

Susanne Bonekamp¹, David Bonekamp¹, Vivek Gowdra Halappa², Jean Francois Geschwind², John Eng², Celia Pamela Corona-Villalobos¹, Diane Reyes², Timothy M. Pawlik³, Ihab R. Kamel²

¹Radiology, Johns Hopkins University, Baltimore, MD, United States; ²Radiology, JHU, Baltimore, MD, United States; ³ Surgery, JHU, Baltimore, MD, United States

One metric that reflects the quality of data collected in clinical studies is the interobserver agreement of a proposed method. In this study we investigated the interobserver agreement of three different approaches used to obtain functional, volumetric and morphologic MRI-based parameters of response to loco-regional treatment in patients with unresectable Hepatocellular Carcinoma. As hypothesized, semi-automated, volumetric analysis of tumor size, tumor enhancement and ADC values resulted in higher interobserver agreement compared to manual ROI-based analysis of the same parameters