(a) An ultrasound image from a localization US dataset of a prostate patient; (b) a 2 MU-AP portal image of fiducial markers; (c) an axial MV-CBCT image of a prostate patient localized with CB, using at isocenter.
Patient mean shift in all three directions for each localization device, along with the overall population mean shift (solid horizontal line) and standard deviation (dashed horizontal lines). One very large patient, whose skin markers were unreliable, was removed from the CBCT data set (open circle).
Scatter plots of the localization data. (a) US AP view; (b) US SI view; (c) US LR view; (d) MV-CBCT AP view; (e) MV-CBCT SI view; (f) MV-CBCT LR view; (g) SM AP view; (h) SM SI view; (i) SM LR view.
Histogram of the mean 3D shift values. (a) US data; (b) MV-CBCT data; (c) SM data.
Mean shift standard deviation in each dimension, for each localization system. The mean 3D shift and its standard deviation are also shown.
Percentage of shifts over in each dimension, for each localization system.
Summary of shifts in literature for US and SM.
P values of the F test for equality of variance between the different pairs of data sets.
values of the Kolmogorov–Smirnov test that a pair of data sets are sampled from the same continuous distribution.
Systematic error , random error and uniform CTV-to-PTV margin suggested by each data set.
Article metrics loading...
Full text loading...