Clinical value of dual energy CT based delineation in the treatment planning for prostate patients
Henriette Klitgaard Mortensen,
Denmark
OC-0466
Abstract
Clinical value of dual energy CT based delineation in the treatment planning for prostate patients
Authors: Henriette Klitgaard Mortensen1, Malene Pilunguaq Rahbek Mathisen2, Janni Mortensen2, Henriette Lindberg3, Lene Sonne Mouritsen4, Line Hammer Dohn3, Jens M. Edmund5
1Herlev and Gentofte Hospital, University of Copenhagen, Department of Oncology, Radiotherapy Unit, 2730 Herlev, Denmark; 2Gentofte and Herlev Hospital, University of Copenhagen, Department of Oncology, Radiotherapy Unit, 2730 Herlev, Denmark; 3Gentofte and Herlev Hospital, University of Copenhagen, Department of Oncology, 2730 Herlev, Denmark; 4Herlev and Gentofte Hospital, University of Copenhagen, Department of Oncology, 2730 Herlev, Denmark; 5Gentofte and Herlev Hospital, University of Copenhagen, Radiotherapy Research Unit, Department of Oncology, 2730 Herlev, Denmark
Show Affiliations
Hide Affiliations
Purpose or Objective
To explore whether selected
dual energy CT (DECT) reconstructions (rec) in addition to standard CT imaging,
would add meaningful support to the delineation process in the radiotherapy (RT)
planning of prostate cancer patients, specifically to benefit those with
contraindication to MR
Material and Methods
We performed a standard
T2-weighted MR imaging on 10 RT patients with prostate cancer. DECT images were
acquired using a dedicated contrast enhanced protocol with a fast 80/140kV
switching DECT technique (GE Revolution CT), with a standard 120kVp-like rec (74keV)
with CT numbers corresponding
to 120kV. 2 additional DECT rec
were further investigated: A 40 keV monochromatic
rec (40keV) and an iodine material density with
water suppression rec (MDIod). All the patients were subject to a combined standard
clinical 74keV/MR delineation process performed by varying oncologists on clinical
duty.
4 structures; prostate,
seminal vesicle (vesicle), rectum and penile bulb delineated as standard combined
74keV/MR imaging, were chosen. These structures were blinded. Three oncologists repeated
the delineation of the structures using only the MDIod, 40keV and 74keV images.
We compared the structure volumes (vol) as measured in Eclipse (Varian Medical
Systems) for prostate and vesicle obtained from 74keV/MR images and the
selected DECT rec. After the delineation, each oncologist evaluated the
clinical usefulness of the 40keV and MDIod rec in an evaluation scheme graded from
0-3 with 0=unusable, 1=slightly usable, 2=usable and 3= very useful. Student-t
tests were carried out to identify significant differences.
Results
Figures 1 and 2 show boxplots
of the 74keV/MR and DECT based structure vol from respectively all patients and
oncologists, of the prostate and vesicles. The mean and standard deviation
(std) of DECT and 74keV/MR based prostate vol were respectively 39,9 +/- 17.7
and 37,6 +/- 15.1 cm3 and non-significantly different (p=0.55). The outliers in
figure 1 were ascribed to one single patient which otherwise indicate a DECT
vol with a smaller interquartile range (IQR) of 16.5 vs. 19.0 for the 74keV/MR
vol.
In figure 2, mean and std
of the vesicles were respectively 9,4 +/- 3,9 and 10,8 +/- 3.2 cm3 for the DECT
and 74keV/MR and non-significantly different (p=0.60). DECT vol were consistently
smaller but with a larger IQR of 5.4 vs. 3.2 for the 74keV/MR.
All 3 oncologists preferred to delineate the structures on the MDIod rec. Grading mean value and std of MDIod were 2.5 (between usable and very useful) +/- 0.65 pooled over all patients and structures. In contrast, the 40KeV grading mean and std was 1.3 (slightly usable) +/- 0.92. There was a significant difference between the usefulness of the two DECT rec (p<1e-16).
Conclusion
DECT based delineation
can be a useful alternative to support the delineation process for prostate
patients undergoing RT planning who cannot participate or have access to MR
imaging.
MDIod reconstruction seems especially useful