Figures & data
Figure 1. DVH of one patient (referred to as Patient 1), for the left LNs with a 5 mm margin. (Left panel) DVHs for single beam proton plans optimized on the pCT and normalized to give the D99% at V99% for the left side of the lymph node volume. (Middle panels) DVHs for the pCT fluence maps applied over the 10 different rCTs. (Right panel) All DVHs in one plot.
Figure 2. Boxplots of V99% for the LNs and V10% for the NT, rectum and bladder for isotropic margins of 0, 3, 5 and 7 mm for all CTs. For each margin and volume, the mean, quartiles, confidence bounds with outliers as well as the minimum and maximum are shown along with the corresponding angle.
Figure 3. Mean variation in WEPL for one of the patients (Patient 1). Angles lower than 55° from the patient axis centered in the prostate were excluded, according to CT limits and physically possible angles. The variation is given in (mm) of water.
Figure 4. WEPL data and DVH data at V99% for the right LNs and at V10% for NT, rectum and bladder. All data normalized to units of standard deviation per dataset, the V% of the LNs are plotted in negative to be comparable on the same scale, meaning the lower value the “better”. At the bottom is the mean of all data, thus in units of mean standard deviation.
Figure 5. Data clustering, at V99% of three patients on the left LNs after log2 transformation, normalization, and selection of most significant angles for p < 0.001 sorted with Pearson's correlation on the y-axis, and all CTs sorted by Pearson's correlation on the x-axis. The branch length in the trees corresponds to the Pearson distance.
