A feasibility study on an automated method to generate patient‐specific dose distributions for radiotherapy using deep learning

Chen, Xinyuan; Men, Kuo; Li, Yexiong; Yi, Junlin; Dai, Jianrong

doi:10.1002/mp.13262

Cited by 154 publications

(177 citation statements)

References 36 publications

Supporting

Mentioning

168

Contrasting

Unclassified

Order By: Relevance

“…The recent evolution of deep learning methods has motivated the use of convolutional neural networks (CNN) to predict patient-specific voxel-wise dose distributions from anatomical information (i.e., contours and/or CT), either in a slice-by-slice manner (2D) [39][40][41]43 or directly as a 3D matrix. 38,42,44 The predicted dose distribution can later be used as an objective to automatically generate a treatment plan.…”

Section: Introductionmentioning

confidence: 99%

Three‐dimensional dose prediction for lung IMRT patients with deep neural networks: robust learning from heterogeneous beam configurations

et al. 2019

View full text Add to dashboard Cite

Purpose The use of neural networks to directly predict three‐dimensional dose distributions for automatic planning is becoming popular. However, the existing methods use only patient anatomy as input and assume consistent beam configuration for all patients in the training database. The purpose of this work was to develop a more general model that considers variable beam configurations in addition to patient anatomy to achieve more comprehensive automatic planning with a potentially easier clinical implementation, without the need to train specific models for different beam settings. Methods The proposed anatomy and beam (AB) model is based on our newly developed deep learning architecture, and hierarchically densely connected U‐Net (HD U‐Net), which combines U‐Net and DenseNet. The AB model contains 10 input channels: one for beam setup and the other 9 for anatomical information (PTV and organs). The beam setup information is represented by a 3D matrix of the non‐modulated beam’s eye view ray‐tracing dose distribution. We used a set of images from 129 patients with lung cancer treated with IMRT with heterogeneous beam configurations (4–9 beams of various orientations) for training/validation (100 patients) and testing (29 patients). Mean squared error was used as the loss function. We evaluated the model’s accuracy by comparing the mean dose, maximum dose, and other relevant dose–volume metrics for the predicted dose distribution against those of the clinically delivered dose distribution. Dice similarity coefficients were computed to address the spatial correspondence of the isodose volumes between the predicted and clinically delivered doses. The model was also compared with our previous work, the anatomy only (AO) model, which does not consider beam setup information and uses only 9 channels for anatomical information. Results The AB model outperformed the AO model, especially in the low and medium dose regions. In terms of dose–volume metrics, AB outperformed AO by about 1–2%. The largest improvement was found to be about 5% in lung volume receiving a dose of 5Gy or more (V5). The improvement for spinal cord maximum dose was also important, that is, 3.6% for cross‐validation and 2.6% for testing. The AB model achieved Dice scores for isodose volumes as much as 10% higher than the AO model in low and medium dose regions and about 2–5% higher in high dose regions. Conclusions The AO model, which does not use beam configuration as input, can still predict dose distributions with reasonable accuracy in high dose regions but introduces large errors in low and medium dose regions for IMRT cases with variable beam numbers and orientations. The proposed AB model outperforms the AO model substantially in low and medium dose regions, and slightly in high dose regions, by considering beam setup information through a cumulative non‐modulated beam’s eye view ray‐tracing dose distribution. This new model represents a major step forward towards predicting 3D dose distributions in real clinical practices, where beam configu...

show abstract

Section: Introductionmentioning

confidence: 99%

Three‐dimensional dose prediction for lung IMRT patients with deep neural networks: robust learning from heterogeneous beam configurations

et al. 2019

View full text Add to dashboard Cite

show abstract

“…The DSC ranged from 0.95 to 1 for different isodose volumes. Chen et al 38 reported for headand-neck cases that the overall mean absolute error for body are 5.5 AE 6.8% and 5.3 AE 6.4% for two types of inputs normalized to prescription. Our results show that the mean absolute 3D dose errors for body is (1.9 AE 1.8)% normalized to prescription.…”

Section: Discussionmentioning

confidence: 99%

“…There are several different types of CNNs, such as LeNet, 33 AlexNet, 24 ZFNet, 34 VGGNet, 35 GoogLeNet, 36 and ResNet, 37 among which ResNet has an excellent performance in ImageNet detection. 37 More recently, Chen et al 38 and Fan et al 39 separately put forward a DL method based on ResNet to predict 3D dose distribution for head-and-neck IMRT patients. Nguyen et al 40 also introduced a U-net 41 model based 3D dose prediction for prostate IMRT patients.…”

Section: Introductionmentioning

confidence: 99%

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

Liu

Fan

et al. 2019

Medical Physics

Self Cite

View full text Add to dashboard Cite

Purpose To develop a deep learning method for prediction of three‐dimensional (3D) voxel‐by‐voxel dose distributions of helical tomotherapy (HT). Methods 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. Results 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. Conclusions 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.

show abstract

“…In addition to KBP, DL has been investigated as a mechanism for automated plan generation (72-75). Fan et al (73) and Chen et al (74) independently undertook the similar task to predict dose distribution with a ResNet DL architecture. Although they trained their models on input data of different types and quantities, both groups demonstrated feasible DL-based automated plans that were similar to expert plans.…”

Section: Treatment Planning and Optimizationmentioning

confidence: 99%

Artificial intelligence in radiation oncology treatment planning: a brief overview

Kiser¹,

Fuller²,

Reed³

2019

J Med Artif Intell

View full text Add to dashboard Cite

Among medical specialties, radiation oncology has long been an innovator and early adopter of therapeutic technologies. This specialty is now situated in prime position to be revolutionized by advances in artificial intelligence (AI), especially machine and deep learning. AI has been investigated by radiation oncologists and physicists in both general and niche radiotherapy planning tasks and has often demonstrated performance that is indistinguishable from human experts, while substantially shortening the time required to complete these tasks. We sought to review applications of AI to domains germane to radiation oncology, namely: image segmentation, treatment plan generation and optimization, normal tissue complication probability modeling, quality assurance (QA), and adaptive re-planning. We sought likewise to consider obstacles to AI adoption in the radiotherapy clinic, now primarily political, legal, and ethical rather than technical in nature.

show abstract

A feasibility study on an automated method to generate patient‐specific dose distributions for radiotherapy using deep learning

Cited by 154 publications

References 36 publications

Three‐dimensional dose prediction for lung IMRT patients with deep neural networks: robust learning from heterogeneous beam configurations

Three‐dimensional dose prediction for lung IMRT patients with deep neural networks: robust learning from heterogeneous beam configurations

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

Artificial intelligence in radiation oncology treatment planning: a brief overview

Contact Info

Product

Resources

About