Meeting Banner
Abstract #0156

Characterising Heterogeneity of Stage 1 Cervical Cancers using Histogram Analysis from Diffusion Weighted Images.

Katherine Downey1,2, S. F. Riches1,2, V. A. Morgan1,2, S. L. Giles1,2, C. Simpkin1,2, D. P. Barton3,4, N. M. deSouza1,2

1Clinical MRI Unit, Institute of Cancer Research, Sutton, United Kingdom; 2Clinical MRI Unit, the Royal Marsden Hospital , Sutton, United Kingdom; 3Gynaecology Unit, the Royal Marsden Hospital, Sutton, United Kingdom; 4Gynaecology Unit, Institute of Cancer Research, Sutton, United Kingdom


This study uses histogram analysis to establish whether ADC values are significantly different in cervical tumours according to their histological characteristics. Patients were scanned with an endovaginal coil to acquire diffusion weighted images. ADC histograms obtained from the entire tumor had 10th, 50th and 90th centile pixel values and skewness of distribution documented. Independent t-test analysis was performed. Increased cellularity of poorly differentiated tumors is reflected in their lower median ADC. ADC distribution does not appear to be associated with lymphovascular space invasion. The increased skewness in the adenocarcinomas is likely to reflect the mixture of glandular and cellular content.