ESTRO 37 Abstract book

S1152

ESTRO 37

1 Kumamoto University Hospital, Radiation Oncology, Kumamoto, Japan 2 Kumamoto University Hospital, Diagnostic Radiology, Kumamoto, Japan 3 Kumamoto University Hospital, Radiological Technology, Kumamoto, Japan 4 Kumamoto University, Medical Imaging, Kumamoto, Japan Purpose or Objective We evaluated the influence of previous treatments on the parametric discrepancies between dose–volume histograms (DVHs) and dose–function histograms (DFHs) generated based on 99 m Tc-labeled diethylene triamine pentaacetate-galactosyl human serum albumin ( 99 m Tc- GSA) single photon emission computed tomography (SPECT) images of hepatocellular carcinoma (HCC) patients treated with stereotactic body radiation therapy (SBRT). Material and Methods This retrospective study based on prospectively acquired data was approved by our institutional review board, and each patient provided written informed consent. Twelve consecutive HCC patients underwent SBRT at 30–40 Gy in five fractions. Surgery had been performed in three (25%) patients before treatment. Nine (75%) patients had undergone radiofrequency ablation, transarterial chemoembolization, and/or percutaneous ethanol injection therapy for HCC in the liver or remnant liver, and no patient received SBRT before treatment. The mean number of previous treatments for lesions other than SBRT-treated lesions was 1.8 (range, 0–4). All patients underwent 99 m Tc-GSA SPECT/ computed tomography (CT) imaging within 1 month before SBRT planning. Attenuation-corrected SPECT and CT images and planning CT images, including delineated structures and dose distributions, were transferred to Velocity AI (version 3.0.2; Varian Medical Systems, Palo Alto, CA, USA). After registration between SPECT and CT images by hardware arrangement, we registered SPECT/CT images onto the planning CT images: a rigid image registration followed by a non-rigid deformable registration. Structures of the irradiated volumes of the liver parenchyma were generated at 5-Gy dose increments, based on the dose distribution information. DFH parameters were calculated as follows: Fx = (sum of the counts within the liver volume receiving a dose of more than x Gy/sum of the counts within the whole liver volume) × 100. The discrepancy between Fx and Vx (Dx = Fx − Vx), where Vx = [(normal liver volume receiving a dose of more than x Gy/whole normal liver volume) × 100], was also calculated. The effect of the number of previous treatments on the absolute value of Dx was evaluated using the Mann–Whitney U test.

Conclusion From the results, RW with a SUV resolution of 0.1 should be recommended as the resampling method to use in PET-TF computation for tumor heterogeneity quantify- cation. EP-2092 Generative Model of Functional RT-Plan Chest CT for Lung Cancer Patients Using Machine Learning B.S. Jang 1 , J.Y. Chang 2 , A.J. Park 3 , H.G. Wu 1 1 Seoul National University Hospital, Radiation Oncology, Seoul, Korea Republic of 2 SMG-SNU Boramae Medical Center, Radiation Oncology, Seoul, Korea Republic of 3 SELVAS AI Incorporation, Artificial Intelligence Research and Development Laboratory, Seoul, Korea Republic of Purpose or Objective Functional image-guided radiotherapy (RT) planning for normal lung avoidance has recently been introduced. SPECT/CT can help identify the functional area of patients, but it is associated with delayed treatment time, additional costs, and unexpected radiation exposure. In this study, we propose a machine learning algorithm that can generate functional chest CT images based on conditional generative adversarial networks (cGAN) using publicly available ‘pix2pix’ implementation. Material and Methods We collected a total of 54 lung perfusion SPECT/CT image sets from lung cancer patients who had been treated at the Seoul National University Hospital. Each SPECT and CT set consists of 161 images. The CT-to-SPECT image pair where there was blank, was outside from the lung parenchymal field, or was not matched anatomically due to the patient's breath was removed under physician’s judgment. After we excluded inappropriate images, 3054 CT-to-SPECT image pair of training sets (49 patients) and 400 of test sets (5 patients) were selected, respectively. Model was trained using ‘pix2pix’ implementation. Results The model was evaluated based on the Multiscale SSIM (MS-SSIM). With 400 image pair of test set, we obtained a lung SPECT/CT fusion image demonstrating MS-SSIM of 0.87 (0.60 – 0.99) compared to original image. This result showed that it could generate functional areas of lung parenchyma directly from chest CT using machine learning algorithm.It suggest that our trained model using cGAN can generate functional areas from RT-plan chest CT images. This could be used for functional image- guided RT planning, sparing patient’s lung function without additional imaging modalities and costs. Further work including planning studies entails with much more training and test set. Conclusion It suggests that our trained model using cGAN can generate functional areas from RT-plan chest CT images. This could be used for functional image-guided RT planning, sparing patient’s lung function without additional imaging modalities and costs. Further work including planning studies entails with much more training and test set. EP-2093 Dose–function histogram evaluation using 99mTc-GSA SPECT/CT images for SBRT planning for HCC R. Toya 1 , T. Saito 1 , S. Shiraishi 2 , Y. Kai 3 , R. Murakami 4 , T. Matsuyama 1 , T. Watakabe 1 , F. Sakamoto 2 , N. Tsuda 2 , Y. Shimohigashi 3 , Y. Yamashita 2 , N. Oya 1

Results Dx %

ranged from −3.4% to 6.8%. Dx was positive for all parameters for five (42%) patients and ranged from negative to positive for seven (58%) patients. The mean