Meeting Banner
Abstract #3438

Measurement Reproducibility of ADC for Liver Metastases Using Multi B Value Diffusion Weighted Imaging: Preliminary Results.

Richard Kinh Gian Do1, David H. Gultekin2, Seth S. Katz1, Michael J. Sohn1, Charles G. Nyman1, Emily C. Zabor3, Chaya M. Moskowitz3, Diane L. Reidy4

1Radiology, Memorial Sloan-Kettering Cancer Center, New York, NY, United States; 2Medical Physics, Memorial Sloan-Kettering Cancer Center, New York, NY, United States; 3Epidemiology and Biostatistics, Memorial Sloan-Kettering Cancer Center, New York, NY, United States; 4Medicine, Memorial Sloan-Kettering Cancer Center, New York, NY, United States

Limited data exist on the reproducibility of ADC of liver metastases using multiple b value DWI. 14 patients with neuroendocrine liver metastases were prospectively recruited. We calculated ADC and ADChigh using 5 b value from 0-500 s/mm2 or 3 b values > 100 s/mm2, respectively. Intra-class correlations (ICC) for ADC of all 44 metastases was 0.845 (CI 0.733,0.912) and for ADChigh was 0.380 (CI: 0.093,0.608). For right hepatic metastases, ADC ICC = 0.895 (CI 0.778,0.952), ADChigh ICC = 0.766 (CI 0.535,0.890). From Bland-Altman analyses, change > 0.56x10-3mm2/s in ADC of metastases (>0.46 x10-3mm2/s for right lobe metastases) would be significant.

Abstract PDF Presentation Video

Keywords

absolute acquisition agreement among apparent applied approved array assess assessment audience biostatistics bland body breath cancer class coefficients coil confidence consecutive considered constraints constructed correlations deviation diffusion encoding epidemiology every examinations excellent exist female fourteen greater hepatic hold improved increasingly indicating initial intensities intervals intra least leaving lesions limited limits linear liver lobe lobes male matrix medical medicine memorial metastases metastasis metastatic monitoring needed oncologists overlap patient patients physicists physics plots poor population potential preliminary promoted prospective radiologist radiologists radiology randomly recorded recruited repeat reproducibility required respectively response scanners selected significantly simultaneously slice slices squares statistical suggest suppressed target therapy treatment true tumor tumors underwent unit versus years