A deep learning model to predict dose-volume histograms of organs at risk in radiotherapy treatment plans.

To develop a deep learning-based model to predict achievable dose-volume histograms (DVHs) of organs at risk (OARs) for automation of inverse planning. The model was based on a connected residual deconvolution network (CResDevNet) and compared with UNet as a baseline. The DVHs of OARs are dependent on patient anatomical features of the planning target volumes and OARs, and these spatial relationships can be learned automatically from prior high-quality plans. The contours of planning target volumes and OARs were parsed from the plan database and used as the input to the model, and the dose-area histograms (DAHs) of the OARs were output from the model. The model was trained from scratch by correlating anatomical features with DAHs of OARs, then accumulating these histograms to obtain the final predicted DVH for each OAR. Helical tomotherapy plans for 170 nasopharyngeal cancer patients were used to train and validate the model. An additional 60 patient treatment plans were used to test the predictive accuracy of the model. The DVHs and dose-volume indices (DVIs) of clinical interest for each OAR in the testing dataset were predicted to evaluate the accuracy of the models. The mean absolute errors in the DVIs for each OAR were calculated using each model and statistically compared using a paired-samples t-test. Dice similarity coefficients for areas of the DVHs were also evaluated. Dose-volume histograms of 21 OARs in nasopharyngeal cancer were predicted using the models. For each patient, 63 DVIs for all OARs were calculated. Using the 60 patient treatment plans in the testing dataset, 79% and 73% of the DVIs predicted using the CResDevNet and UNet models, respectively, were within 5% of the clinical values. The median value of the DVIs' mean absolute errors was 3.2 ± 2.5% and 3.7 ± 2.9% for the CResDevNet and UNet models, respectively. The average dice similarity coefficient for all OARs was 0.965 using the CResDevNet model and 0.958 using the UNet model. A deep learning model was developed for predicting achievable DVHs of OARs. The prediction accuracy of the CResDevNet model was evaluated using a planning database of nasopharyngeal cancer cases and shown to be more accurate than the UNet model. Prediction accuracy was also higher for larger-volume OARs. The model can be used for automation of inverse planning and quality assessment of individual treatment plans.

Liu Z ，Chen X ，Men K ，Yi J ，Dai J ... -  《-》

Quantitative analysis of the factors which affect the interpatient organ-at-risk dose sparing variation in IMRT plans.

Yuan L ，Ge Y ，Lee WR ，Yin FF ，Kirkpatrick JP ，Wu QJ ... -  《MEDICAL PHYSICS》

DVHnet: A deep learning-based prediction of patient-specific dose volume histograms for radiotherapy planning.

To develop a deep learning method to predict patient-specific dose volume histograms (DVHs) for radiotherapy planning. Patient data included 180 cases with nasopharyngeal cancer, of which 153 cases were used for training and 27 for testing. A network (named "DVHnet") based on a convolutional neural network (CNN) was designed for directly predicting DVHs of organs at risk (OARs). Two-channel images with contoured structures were generated as the inputs for training the model. A one-dimensional array consisting of 256 continuous volume percentages on a DVH curve for each slice was calculated as the corresponding output. The combined DVH was then calculated. Sixteen OARs were modeled in the study. Prediction accuracy was evaluated against the corresponding DVH curve of ground truth (GT) plans. A global DVH analysis and critical dosimetry metrics for each OAR were calculated for quantitative evaluation. The performance of DVHnet also was evaluated against two baselines: DosemapNet (developed by our research group) and commercial RapidPlan software. The predicted mean difference in average dose of all OARs using DVHnet was 0.30 ± 0.95 Gy. And the predicted differences in D2% and D50 can be control within 2.32 and 0.69 Gy. For most OARs, there were no obvious differences between the dosimetric metrics of the predicted and GT values for both DVHnet and DosemapNet (P ≥ 0.05). Only the predicted D2% of the optic organs for DVHnet, and of brain stem PRV for DosemapNet displayed statistically significant differences. Except for the optic organs, DVHnet performs better than or comparably with RapidPlan. The mean difference in proportion of points of interest was 3.59% ± 7.78%. A deep learning network model was developed to automatically extract useful features for accurate prediction of patient-specific DVH curves directly. The performance of DVHnet was comparable to DosemapNet and RapidPlan.

Chen X ，Men K ，Zhu J ，Yang B ，Li M ，Liu Z ，Yan X ，Yi J ，Dai J ... -  《-》

A deep learning method for prediction of three-dimensional dose distribution of helical tomotherapy.

To develop a deep learning method for prediction of three-dimensional (3D) voxel-by-voxel dose distributions of helical tomotherapy (HT). Using previously treated HT plans as training data, a deep learning model named U-ResNet-D was trained to predict a 3D dose distribution. First, the contoured structures and dose volumes were converted from plan database to 3D matrix with a program based on a developed visualization toolkit (VTK), then transferred to U-ResNet-D for correlating anatomical features and dose distributions at voxel-level. One hundred and ninety nasopharyngeal cancer (NPC) patients treated by HT with multiple planning target volumes (PTVs) in different prescription patterns were studied. The model was typically trained from scratch with weights randomly initialized rather than using transfer-learning method, and used to predict new patient's 3D dose distributions. The predictive accuracy was evaluated with three methods: (a) The dose difference at the position r, δ(r, r) = Dc (r) - Dp (r), was calculated for each voxel. The mean (μδ(r,r) ) and standard deviation (σδ(r,r) ) of δ(r, r) were calculated to assess the prediction bias and precision; (b) The mean absolute differences of dosimetric indexes (DIs) including maximum and mean dose, homogeneity index, conformity index, and dose spillage for PTVs and organ at risks (OARs) were calculated and statistically analyzed with the paired-samples t test; (c) Dice similarity coefficients (DSC) between predicted and clinical isodose volumes were calculated. The U-ResNet-D model predicted 3D dose distribution accurately. For twenty tested patients, the prediction bias ranged from -2.0% to 2.3% and prediction error varied from 1.5% to 4.5% (relative to prescription) for 3D dose differences. The mean absolute dose differences for PTVs and OARs are within 2.0% and 4.2%, and nearly all the DIs for PTVs and OARs had no significant differences. The averaged DSC ranged from 0.95 to 1 for different isodose volumes. The study developed a new deep learning method for 3D voxel-by-voxel dose prediction, and shown to be able to produce accurately dose predictions for nasopharyngeal patients treated by HT. The predicted 3D dose map can be useful for improving radiotherapy planning design, ensuring plan quality and consistency, making clinical technique comparison, and guiding automatic treatment planning.

Liu Z ，Fan J ，Li M ，Yan H ，Hu Z ，Huang P ，Tian Y ，Miao J ，Dai J ... -  《-》

Evaluation of dose-volume histogram prediction for organ-at risk and planning target volume based on machine learning.

Jiao SX ，Wang ML ，Chen LX ，Liu XW ... -  《Scientific Reports》