Meeting Banner
Abstract #4149

The Effects of Region of Interest Errors on Estimates of Whole-Body Tumour Volume and ADC in Patients with Bone Metastases.

Matthew D. Blackledge1, Nina Tunariu1, David John Collins1, Anwar R. Padhani2, Martin O. Leach1, Dow M. Koh1

1CRUK and EPSRC Cancer Imaging Centre, Institute of Cancer Research, Sutton, Surrey, United Kingdom; 2Paul Strickland Scanner Centre, Mount Vernon Hospital, Northwood, Middlesex, United Kingdom

We derive the intra-observer repeatability (Bland-Altman statistics) of two radiologists reporting median ADC and total diffusion volume estimates derived from whole-body diffusion-weighted imaging (WBDWI) in a cohort of nine patients with bone metastases. Good repeatability in the range -13.5 to 15.6% is found for ADC estimates and -44.2% to 79.3% for volume estimates. Clinical changes of these parameters in response to treatment are shown to be greater than the repeatability limits observed, as demonstrated in a pilot cohort of 13 patients with bone metastases.

Keywords

achieved acquisition agreement alongside alternatives although anatomical annual apparent applied assessing authors automatic background biomedical bland blinded body bone breast breathing cancer cancers clinical clinically cohort column columns concluded correct corrected correction covering criteria defined defining definition delineation derive derived detecting developing diagnosed diagnostic diffusion disease error errors especially evaluating exclusion experience feasibility field free funding furthermore generated global good grant greater health highlights importance improved include included inclusion independent indicating inspection inter intersection interval intra investigate kingdom knowledge leach lesions limits location malignancies manually maps marrow median medical meeting metastases metastatic metrics milliliters model mono month morphological mount necrosis nine observer observers outcomes outlined patient patients percentage pilot pixel population practice primary prostate quantitative radiologist radiologists reading receiver removal repeatability report reporting represent respectively response revealed review scanner score segmentation semi seven skull stations statistician stir substantially suppressed suppression surrey system table thigh throughout treatment twice vital volume volumes whole