ESTRO 2023 - Abstract Book

S502

Sunday 14 May 2023

ESTRO 2023

Figure 2: Cumulative (a) physical and (b) EQD2 cumulative dose metrics for the head and neck (a & c) and lung cancer (b & d) cases reported by the participants (n=21) using their clinical pathway. (e) & (f) Cumulative physical dose metrics for doses summed and reported in a single treatment planning software (TPS) for the HNC and lung cancer cases respectively, based on mapped dose distributions. Conclusion Differences in methods for spatial mapping of dose between courses, conversion to biologically equivalent doses and use of tissue recovery factors resulted in substantial variations in cumulative dose assessment in two re-irradiation cases. The variations observed have implications for outcome analysis and our understanding of published doses concerning re- irradiation. Standardising the workflow using spatially registered doses may present a pathway for improved consistency in cumulative dose assessment. OC-0616 Generative adversarial networks for head-and-neck cancer radiotherapy dose distribution prediction V. Strijbis 1 , X. Gu 1 , M. Dahele 1 , B. Slotman 1 , W. Verbakel 1 1 Amsterdam UMC, Cancer Center Amsterdam, Radiation Oncology, Amsterdam, The Netherlands Purpose or Objective Head-and-neck cancer (HNC) radiotherapy is complex, requires contouring of multiple target volumes and a large range of organs-at-risk (OARs) and is followed by a time-consuming treatment planning process to obtain a near-optimal dose distributions. Because of their focus on realism through adversarial learning, generative adversarial networks (GAN) are suitable for dose distribution prediction from a treatment planning perspective. In addition, if the input data required for dose prediction can be limited, treatment planning efficiency may be further improved. Motivated by this, we explored the relationship between various combinations of input data and how well the GAN-predicted dose distribution agreed with the benchmark clinical plan. Materials and Methods Data from 355 HNC patients (300/20/35 train/validation/test) treated between 2013-2019 were used: (1) planning CT, (2) OAR contours (salivary glands, swallowing structures, oral cavity, spinal cord), (3) PTV structures (elective, boost), (4) body contour, (5) clinical dose distribution. CT scans were window-levelled from -300 to +200 HU, cropped and down-sampled by half to a 128x128x64 grid with a resolution of 2x2x5mm3. Generator and discriminator networks used UNet and VGG- type architectures, respectively. The GAN-loss was a linear combination of binary cross-entropy and L1 and L2 reconstruction loss. GAN dose distributions were compared to clinical dose using dose volume histograms and mean OAR and PTV dose differences with clinical plans. We investigated input combinations: (1) CT, (2) CT+PTV, (3) CT+PTV+OAR, (4) PTV+OAR, (5) PTV+body. Mann-Whitney U-tests were used to assess statistical significance. Results For each respective input combination, median[IQR] parotid L1-losses were 6.9[3.9], 2.8[0.6], 2.4[1.2], 2.4[1.5], 3.1[1.7]Gy. CT+PTV+OAR resulted in the lowest median[IQR] L1-loss for the entire body (3.3[0.6]Gy) and was significantly better than CT+PTV and PTV+body (p=0.044, 0.002, respectively).No significant differences were found between CT+PTV & PTV+body, and CT+PTV+OAR & PTV+OAR. CT-only resulted in considerably and significantly (p<0.001) worse dose distributions compared to all other input combinations.