ESTRO 2025 - Abstract Book

S3755

Physics - Radiomics, functional and biological imaging and outcome prediction

ESTRO 2025

1732

Mini-Oral Feasibility Study of Deep Learning-Based MRI-to-PET Generation for Rectal Cancer: Overall Survival and Pathological Complete Response Assessment Xiaojie Yin, Jiazhou Wang, Weigang Hu, Zhen Zhang, Zhenhao Li, Yue Zou, Ziwei Li Department of Radiation Oncology, Fudan University Shanghai Cancer Center, Shanghai, China Purpose/Objective: This study aims to develop and validate a novel deep learning method to generate synthetic PET images for rectal cancer from MRI data. By incorporating metabolic information from the synthetic PET images, we seek to improve tumor assessment, offering new approaches for accurate diagnosis and prognosis prediction in rectal cancer. Material/Methods: A deep learning model was developed to synthesize PET images from MRI data. The model was trained on data from 150 patients with locally advanced rectal cancer, using baseline MR and PET/CT images acquired within one month. Model performance was evaluated by comparing features of synthetic and real PET images from 21 patients, focusing on key metrics such as metabolic tumor volume at SUV thresholds of 4 (MTV 4 ) and 6 (MTV 6 ), as well as SUV max and SUV mean . Two additional datasets, comprising 392 and 346 patients, were used to assess the clinical value of the synthetic PET images in overall survival (OS) prediction and pathological complete response (pCR) evaluation. For pCR evaluation, synthetic PET images from both before and after neoadjuvant chemoradiotherapy (nCRT) were integrated, along with clinical indicators, to develop a comprehensive validation model. Additionally, an external validation was conducted using PET/MR images from three rectal cancer patients from another institution. Results: The correlation coefficients between the synthetic and real PET images for MTV 4 and MTV 6 were 0.82 and 0.75, respectively. In terms of OS prediction, indicators such as MTV 4 and MTV 6 emerged as independent prognostic factors, with a hazard ratio (HR) of 1.76 (95% CI: 1.24-2.51) for MTV 4 (p=0.004) and 1.56 (95% CI: 1.09-2.21) for MTV 6 (p=0.032). The AUC for predicting pCR based on synthetic PET images before nCRT was 0.69, which increased to 0.84 when combined with other clinical indicators and post-nCRT imaging. In the external validation, the SSIM and PSNR of the MR-synthetic PET compared to concurrently acquired PET images were 0.79 and 24.59 dB, respectively.