AUTHOR=Gao Yuan , Chang Chih-Wei , Roper Justin , Axente Marian , Lei Yang , Pan Shaoyan , Bradley Jeffrey D. , Zhou Jun , Liu Tian , Yang Xiaofeng TITLE=Single energy CT-based mass density and relative stopping power estimation for proton therapy using deep learning method JOURNAL=Frontiers in Oncology VOLUME=Volume 13 - 2023 YEAR=2023 URL=https://www.frontiersin.org/journals/oncology/articles/10.3389/fonc.2023.1278180 DOI=10.3389/fonc.2023.1278180 ISSN=2234-943X ABSTRACT=The number of patients undergoing proton therapy has increased in recent years. Current treatment planning systems (TPS) calculate dose map using three-dimensional (3D) maps of relative stopping power (RSP) and mass density. The patient specific maps of RSP and mass density were obtained from translating CT number (HU) acquired using single-energy computed tomography (SECT) with appropriate conversions and coefficients. SECT is the major clinical modality for proton therapy treatment planning. It would be intriguing to enhance proton dose calculation accuracy using a deep learning (DL) approach centered on SECT.The purpose of this work is to develop a deep learning method to generate mass density and relative stopping power (RSP) maps based on clinical single-energy CT (SECT) data for proton dose calculation in proton therapy treatment.Methods: Artificial neural networks (ANN), fully convolutional neutral networks (FCNN) and residual neural networks (ResNet) were used to learn the correlation between voxel specific mass density, RSP and SECT CT number (HU). Stoichiometric calibration method based on SECT data and an empirical model based on dual-energy CT (DECT) images were chosen as reference models to evaluate the performance of deep learning neural networks. SECT images of a CIRS 062M electron density phantom was used as the training dataset for deep learning models. CIRS anthropomorphic M701 and M702 phantoms were used to test the performance of deep learning models.Results: For M701, the mean absolute percentage errors (MAPE) of the mass density map by FCNN are 0.39%, 0.92%, 0.68%, 0.92%, and 1.57% on brain, spinal cord, soft tissue, bone, and lung, compared to that of SECT stoichiometric method are 0.99%, 2.34%, 1.87%, 2.90% and 12.96%. For RSP maps, MAPE of FCNN on M701 are 0.85%, 2.32%, 0.75%, 1.22%, 1.25%, compared with that of the SECT reference model are 0.95%, 2.61%, 2.08%, 7.74% and 8.62%.The results show that the deep learning neural networks have the potential to generate accurate voxel specific material property information, which can be used to improve the accuracy of proton dose calculation.Deep learning-based frameworks are proposed to estimate material mass density and RSP from SECT with improved accuracy compared with conventional method.