820 / 1119
systematic underestimation of rectum D
2cm³
:
rectum probe (iv. dosimetry) stays inside -
rectum always fills with gas in between image
acquisition and treatment -> D
2cm³
is
systematically higher for each fraction
Systematic dosimetric uncertainties
small targets,
good coverage
large targets,
less coverage
Systematic inter-/intra fraction variations for MRI-
based cervix BT
(Nesvacil et al. 2013, R&O 107):
e.g.
∆
D90< +3%/fx => „observed“ local
control @85 Gy 1% higher than model
prediction
Systematically larger contours on CT vs. MRI =>
underestimation of D90 by CT contours
(e.g.
Viswanathan et al. 2007, IJROBP 68):
e.g. i)
∆
D
90
=+10%/fx => 2% overestimation of local
control @ 85 Gy
ii)
∆
D
90
=+20%/fx => 3.5% overestimation
of local control @ 85 Gy
“Can reduction of uncertainties in cervix cancer brachytherapy potentially
improve clinical outcome?” Nesvacil et al. 2016, submitted to R&O
e.g. i)
∆
D
2cm³
=+3%/fx => observed
morbidity @75Gy 1% higher than
model prediction
ii)
∆
D
2cm³
=+5%/fx => observed morbidity
@75 Gy is 2% higher than model
prediction