Cone‐beam CT‐derived relative stopping power map generation via deep learning for proton radiotherapy

Abstract

In intensity-modulated proton therapy (IMPT), protons are used to deliver highly conformal dose distributions, targeting tumors, and sparing organs-at-risk. However, due to uncertainties in both patient setup and relative stopping power (RSP) calculation, margins are added to the treatment volume during treatment planning, leading to higher doses to normal tissues. Cone-beam computed tomography (CBCT) images are taken daily before treatment; however, the poor image quality of CBCT limits the use of these images for online dose calculation. In this work, we use a deep-learning-based method to predict RSP maps from daily CBCT images, allowing for online dose calculation in a step toward adaptive radiation therapy. Methods: Twenty-three head-and-neck cancer patients were simulated using a Siemens TwinBeam dual-energy CT (DECT) scanner. Mixed-energy scans (equivalent to a 120 kVp single-energy CT scan) were converted to RSP maps for treatment planning. Cone-beam computed tomography images were taken on the first day of treatment, and the planning RSP maps were registered to these images. A deep learning network based on a cycle-GAN architecture, relying on a compound loss function designed for structural and contrast preservation, was then trained to create an RSP map from a CBCT image. Leave-one-out and holdout cross validations were used for evaluation, and mean absolute error (MAE), mean error (ME), peak signal-to-noise ratio (PSNR), and structural similarity (SSIM) were used to quantify the differences between the CT-based and CBCT-based RSP maps. The proposed method was compared to a deformable image registration-based method which was taken as the ground truth and two other deep learning methods. For one patient who underwent resimulation, the new planning RSP maps and CBCT images were used for further evaluation and validation. Results: The CBCT-based RSP generation method was evaluated on 23 head-and-neck cancer patients. From leave-one-out testing, the MAE between CT-based and CBCT-based RSP was 0.06 AE 0.01 and the ME was −0.01 AE 0.01. The proposed method statistically outperformed the comparison DL methods in terms of MAE and ME when compared to the planning CT. In terms of dose comparison, the mean gamma passing rate at 3%/3 mm was 94% when three-dimensional (3D) gamma index was calculated per plan and 96% when gamma index was calculated per field. Conclusions: The proposed method provides sufficiently accurate RSP map generation from CBCT images, allowing for evaluation of daily dose based on CBCT and possibly allowing for CBCTguided adaptive treatment planning for IMPT.

Deep learning based synthetic‐CT generation in radiotherapy and PET: A review

“…For proton RT, the setup accuracy and dose calculation are even more relevant to avoid RS errors that could jeopardise the benefit of treatment 63 . Because there is an intrinsic error in converting HU to relative proton stopping power, 178 it has been shown that DL methods can translate CBCT directly to stopping power 179 . This approach has not been covered in this review, but it is an exciting approach that will probably lead to further investigations.Interestingly, increasing the quality of CBCT can be tackled not only as an image‐to‐image translation problem but also as an inverse problem, that is, from a reconstruction perspective.…”

Dosimetric evaluation of cone‐beam CT‐based synthetic CTs in pediatric patients undergoing intensity‐modulated proton therapy

“…This is consistent with what has been reported in adult head and neck patients (average passing rate of 94% for 3%/3 mm) who were originally treated with photons and were retrospectively replanned for proton therapy. 19 It is interesting to note that the sites with the lowest pass rates corresponded to the mediastinum (87%) and lung (90%). This may reflect the limitations of deformable registration, where the displacement between ribs and lung may not be accurately modeled without specialized algorithms.…”

Management of Motion and Anatomical Variations in Charged Particle Therapy: Past, Present, and Into the Future

“…presented a single detector, multi-energy proton radiography system which relied on artificial intelligence to filter out proton scatter ( 198 ). In addition, multiple studies have been reported which utilize deep learning as a tool to facilitate accurate proton dose calculation from daily CBCT images ( 121 , 199 – 201 ). Elmahdy et al.…”

“…Van der Heyden et al presented a single detector, multi-energy proton radiography system which relied on artificial intelligence to filter out proton scatter (198). In addition, multiple studies have been reported which utilize deep learning as a tool to facilitate accurate proton dose calculation from daily CBCT images (121,(199)(200)(201). Elmahdy et al also reported on a convolutional neural network (CNN)-based model for robust, automatic contour propagation in prostate cancer for online adaptive proton therapy (202).…”