Comparison of CTVHR and organs at risk contours between TRUS and MR images in IB cervical cancers: a proof of concept study

Comparison of CTVHR and organs at risk contours between TRUS and MR images in IB cervical cancers: a proof of concept study

Abstract Purpose To compare CTVHR and OAR dimensions and inter-rater agreement between magnetic resonance (MR) and trans-rectal ultrasound (TRUS) images in IB cervical cancer patients. Methods IB cervical cancer patients treated with (chemo)radiotherapy plus MR-guided brachytherapy (BT) were prospec...

Full description

Bibliographic Details
Main Authors: Lucas C. Mendez, Ananth Ravi, Kevin Martell, Hamid Raziee, Yasir Alayed, Matt Wronski, Moti Paudel, Elizabeth Barnes, Amandeep Taggar, C. S. Wong, Eric Leung
Format: Article
Language:English
Published: BMC 2020-04-01
Series:Radiation Oncology
Subjects:
Cervical cancer
Brachytherapy
Transrectal ultrasound
Online Access:http://link.springer.com/article/10.1186/s13014-020-01516-4
Description
Summary:Abstract Purpose To compare CTVHR and OAR dimensions and inter-rater agreement between magnetic resonance (MR) and trans-rectal ultrasound (TRUS) images in IB cervical cancer patients. Methods IB cervical cancer patients treated with (chemo)radiotherapy plus MR-guided brachytherapy (BT) were prospectively enrolled in this study. Radiation oncologists contoured CTVHR and OARs in pre-BT MR images (MRI) and intra-operative TRUS images. These contours were subsequently compared in regard to volume and dimension. Contour inter-rater agreement analysis was also investigated using kappa index (KI). Stata 15.0 was used for statistical analysis and a p-value < 0.05 was considered statistically significant. Results TRUS CTVHR volumes were statistically smaller than the respective MRI contoured volumes. TRUS CTVHR thickness was found to be consistently smaller than MRI contours in all patients. No statistical difference was seen in width and height between the two different imaging modalities. MRI contours had a median KI of 0.66 (range: 0.56–0.77) while TRUS-based contours had a median KI of 0.64 (range: 0.47–0.77). Bladder and rectum had very satisfactory KI in both imaging modalities. Vaginal contours had moderate agreement in MR (0.52) and in TRUS images (0.58). Conclusion TRUS images allow good visualization of CTVHR and OARs in IB cervical cancer patients. Inter-rater contour variability was comparable between TRUS and MR images. TRUS is a promising modality on its own for image-guided BT.
ISSN:1748-717X

Similar Items