Boosting radiotherapy dose calculation accuracy with deep learning

Abstract

In radiotherapy, a trade-off exists between computational workload/speed and dose calculation accuracy. Calculation methods like pencil-beam convolution can be much faster than Monte-Carlo methods, but less accurate. The dose difference, mostly caused by inhomogeneities and electronic disequilibrium, is highly correlated with the dose distribution and the underlying anatomical tissue density. We hypothesize that a conversion scheme can be established to boost low-accuracy doses to highaccuracy, using intensity information obtained from computed tomography (CT) images. A deep learning-driven framework was developed to test the hypothesis by converting between two commercially available dose calculation methods: Anisotropic analytic algorithm (AAA) and Acuros XB (AXB). A hierarchically dense U-Net model was developed to boost the accuracy of AAA dose toward the AXB level. The network contained multiple layers of varying feature sizes to learn their dose differences, in relationship to CT, both locally and globally. Anisotropic analytic algorithm and AXB doses were calculated in pairs for 120 lung radiotherapy plans covering various treatment techniques, beam energies, tumor locations, and dose levels. For each case, the CT and the AAA dose were used as the input and the AXB dose as the "ground-truth" output, to train and test the model. The mean squared errors (MSEs) and gamma passing rates (2 mm/2% & 1 mm/1%) were calculated between the boosted AAA doses and the "ground-truth" AXB doses. The boosted AAA doses demonstrated substantially improved match to the "ground-truth" AXB doses, with average (± s.d.) gamma passing rate (1 mm/1%) 97.6% (±2.4%) compared to 87.8% (±9.0%) of the original AAA doses. The corresponding average MSE was 0.11(±0.05) vs 0.31(±0.21). Deep learning is able to capture the differences between dose calculation algorithms to boost the low-accuracy algorithms. By combining a less accurate dose calculation algorithm with a trained deep learning model, dose calculation can potentially achieve both high accuracy and efficiency.

Artificial intelligence and machine learning for medical imaging: A technology review

“…For instance, instead of using generic loss functions from computer vision tasks, like the mean squared error, one could use loss functions that better target the specificities of our medical problem, such as including mutual information for the conversion of different image modalities [90,193] or dose-volume histograms for radiotherapy dose predictions [212]. Regarding the injection of domain-specific knowledge as input to the models, some examples include the addition of electronic health records and clinical data, like text and laboratory results, to the image data [213][214][215], or having first-order prior or approximations of the expected output [175,[216][217][218][219] Integrating domain-specific knowledge cannot only serve to improve the performances of state-of-the-art AI models, but also to increase the interpretability of the results, which is one of the well-acknowledged limitations of the current ML/DL methods [220][221][222][223]. This is the idea behind the so-called Expert Augmented Machine Learning (EAML), whose goal is to develop algorithms capable of extracting human knowledge from a panel of experts and use it to establish constraints for the model's prediction [224].…”

NeuralDAO: Incorporating neural network generated dose into direct aperture optimization for end‐to‐end IMRT planning

“…For reference, precalculating the dose deposition matrix with PB takes 30 s. The added overhead is acceptable. It may be computationally efficient to produce the accurate dose without the open‐field MC dose channel, as evidenced by other works 16–19 . However, enabling this channel leads to faster convergence, as observed in our experiment.…”

“…Following the previous research's success, we design a neural dose network f neu that produces the highaccuracy unit neural dose dneu k from the low-accuracy unit PB dose dPB k for an aperture shape x k : [16][17][18][19] However, enabling this channel leads to faster convergence, as observed in our experiment.…”

Deep learning dose prediction for IMRT of esophageal cancer: The effect of data quality and quantity on model performance

“…The DL model architecture was inspired from the popular U-Net [36], a type of fully convolutional network widely used for medical image segmentation and other medical applications [19,[51][52][53]. U-Net type of architectures have gained interest for the prediction of dose distributions in radiotherapy treatments, due to its ability to include both local and global features from the input images (i.e.…”