MR‐based treatment planning in radiation therapy using a deep learning approach

Abstract

Purpose To develop and evaluate the feasibility of deep learning approaches for MR‐based treatment planning (deepMTP) in brain tumor radiation therapy. Methods and materials A treatment planning pipeline was constructed using a deep learning approach to generate continuously valued pseudo CT images from MR images. A deep convolutional neural network was designed to identify tissue features in volumetric head MR images training with co‐registered kVCT images. A set of 40 retrospective 3D T1‐weighted head images was utilized to train the model, and evaluated in 10 clinical cases with brain metastases by comparing treatment plans using deep learning generated pseudo CT and using an acquired planning kVCT. Paired‐sample Wilcoxon signed rank sum tests were used for statistical analysis to compare dosimetric parameters of plans made with pseudo CT images generated from deepMTP to those made with kVCT‐based clinical treatment plan (CTTP). Results deepMTP provides an accurate pseudo CT with Dice coefficients for air: 0.95 ± 0.01, soft tissue: 0.94 ± 0.02, and bone: 0.85 ± 0.02 and a mean absolute error of 75 ± 23 HU compared with acquired kVCTs. The absolute percentage differences of dosimetric parameters between deepMTP and CTTP was 0.24% ± 0.46% for planning target volume (PTV) volume, 1.39% ± 1.31% for maximum dose and 0.27% ± 0.79% for the PTV receiving 95% of the prescribed dose (V95). Furthermore, no significant difference was found for PTV volume (P = 0.50), the maximum dose (P = 0.83) and V95 (P = 0.19) between deepMTP and CTTP. Conclusions We have developed an automated approach (deepMTP) that allows generation of a continuously valued pseudo CT from a single high‐resolution 3D MR image and evaluated it in partial brain tumor treatment planning. The deepMTP provided dose distribution with no significant difference relative to a kVCT‐based standard volumetric modulated arc therapy plans.

Dosimetric evaluation of synthetic CT image generated using a neural network for MR‐only brain radiotherapy

“…The average of MAE between synthetic CT and planning CT was 60.77 AE 13.99 HU, which is better than some previously reported results using atlas-based or machine learning methods. [25][26][27][28][29][30][31] A few studies used the neural network, 12,13,16,17,32 either GAN or CNN.…”

“…Since our prior goal was to use the synthetic CT from MRI for dose calculation in clinical workflow, the dosimetric performance of synthetic CT was also evaluated by comparing dose distribution of the same clinical plan delivered, respectively, on synthetic CT and planning CT. A mean pass rate of 99.79% with 3 mm/ 3 % criterion and 97.23% with 2 mm/2% criterion was similar to other studies using deep learning approaches. 13,33 While Kazemifar 17 also utilized GAN network and show a slightly higher pass rate of 98.7% with 2 mm/2%. One of the distinctions in their study was the use of mutual information as the loss function, instead of MAE in ours.…”

“…Several studies utilized the convolutional neural network to perform the synthesis of CT from a variety of MRI sequences. Han 12 and Liu 13 applied the u‐net 14 based network to convert MRI to sCT pixel by pixel. The encoder‐decoder architecture in their networks enable the learning of a hierarchy of features from MRI through a downsampling process, then those features in various resolution were combined to generate high‐resolution CT image through an upsampling process.…”

“…For all testing patients, the average and standard deviation of MAE between synthetic and real CT were 60 13. AE 13.72 HU, 60.51 AE 14.23 HU, 61.14 AE 12.56 HU, 62.28 AE 13.85 HU, 59.38 AE 13.23 HU using training folds 1, 2, 3, 4, and 5, respectively.…”

Adapting training for medical physicists to match future trends in radiation oncology

“…These include automation of treatment planning [31] , adaptive radiotherapy, MR-linac systems [32] , biological and functional imaging [33] , dose painting [34] , radiomics [35] , dosiomics [36] , and predictive modelling [37] . There is also a wide range of topics investigated with artificial intelligence (neural networks, deep learning [38] , [39] ), including segmentation of tumors and OARs [40] , pseudo-CT generation from MRI [41] , dose prediction for treatment planning [42] , patient-specific quality assurance [43] , real-time respiratory motion prediction [44] , and prediction of treatment response [45] .…”