Improved contrast and noise of megavoltage computed tomography (MVCT) through cycle-consistent generative machine learning.

PURPOSE Megavoltage computed tomography (MVCT) has been implemented on many radiation therapy treatment machines as a tomographic imaging modality that allows for three-dimensional visualization and localization of patient anatomy. Yet MVCT images exhibit lower contrast and greater noise than its kilovoltage CT (kVCT) counterpart. In this work, we sought to improve these disadvantages of MVCT images through an image-to-image based machine learning transformation of MVCT and kVCT images. We demonstrated that by learning the style of kVCT images, MVCT images can be converted into high quality synthetic kVCT (skVCT) images with higher contrast and lower noise, when compared to the original MVCT. METHODS kVCT and MVCT images of 120 head and neck (H&N) cancer patients treated on an Accuray TomoHD system were retrospectively analyzed in this study. A cycle-consistent generative adversarial network (CycleGAN) machine learning, a variant of the generative adversarial network (GAN), was used to learn Hounsfield Unit (HU) transformations from MVCT to kVCT images, creating skVCT images. A formal mathematical proof is given describing the interplay between function sensitivity and input noise and how it applies to the error variance of a high-capacity function trained with noisy input data. Finally, we show how skVCT shares distributional similarity to kVCT for various macro-structures found in the body. RESULTS Signal to noise ratio (SNR) and contrast to noise ratio (CNR) were improved in skVCT images relative to the original MVCT images and were consistent with kVCT images. Specifically, skVCT CNR for muscle - fat, bone - fat and bone - muscle improved to 14.8 ± 0.4, 122.7 ± 22.6, and 107.9 ± 22.4 compared with 1.6 ± 0.3, 7.6 ± 1.9, and 6.0 ± 1.7, respectively, in the original MVCT images and was more consistent with kVCT CNR values of 15.2 ± 0.8, 124.9 ± 27.0, and 109.7 ± 26.5, respectively. Noise was significantly reduced in skVCT images with SNR values improving by roughly an order of magnitude and consistent with kVCT SNR values. Axial slice mean (S-ME) and mean absolute error (S-MAE) agreement between kVCT and MVCT/skVCT improved, on average, from -16.0 and 109.1 HU to 8.4 and 76.9 HU, respectively. CONCLUSIONS A kVCT-like qualitative aid was generated from input MVCT data through a CycleGAN instance. This qualitative aid, skVCT, was robust towards embedded metallic material, dramatically improves HU alignment from MVCT, and appears perceptually similar to kVCT with SNR and CNR values equivalent to that of kVCT images.

[1]  Timothy D. Solberg,et al.  Attention-Aware Discrimination for MR-to-CT Image Translation Using Cycle-Consistent Generative Adversarial Networks. , 2020, Radiology. Artificial intelligence.

[2]  Boaz Barak,et al.  Deep double descent: where bigger models and more data hurt , 2019, ICLR.

[3]  Timothy D Solberg,et al.  Attention-enabled 3D boosted convolutional neural networks for semantic CT segmentation using deep supervision , 2019, Physics in medicine and biology.

[4]  Jiwen Dong,et al.  A Megavoltage CT Image Enhancement Method for Image-Guided and Adaptive Helical TomoTherapy , 2019, Front. Oncol..

[5]  Vasant Kearney,et al.  A convolutional neural network algorithm for automatic segmentation of head and neck organs at risk using deep lifelong learning. , 2019, Medical physics.

[6]  S. Kida,et al.  Visual enhancement of Cone-beam CT by use of CycleGAN. , 2019, Medical physics.

[7]  Vasant Kearney,et al.  DoseNet: a volumetric dose prediction algorithm using 3D fully-convolutional neural networks , 2018, Physics in medicine and biology.

[8]  T. Solberg,et al.  A Multi-Task CNN Model for Autosegmentation of Prostate Patients , 2018, International Journal of Radiation Oncology*Biology*Physics.

[9]  T. Solberg,et al.  Deep Learning Misuse in Radiation Oncology , 2018, International Journal of Radiation Oncology*Biology*Physics.

[10]  Jing Wang,et al.  Generating synthesized computed tomography (CT) from cone-beam computed tomography (CBCT) using CycleGAN for adaptive radiation therapy , 2018, Physics in medicine and biology.

[11]  Gilmer Valdes,et al.  An unsupervised convolutional neural network-based algorithm for deformable image registration , 2018, Physics in medicine and biology.

[12]  Kun Zhang,et al.  Geometry-Consistent Generative Adversarial Networks for One-Sided Unsupervised Domain Mapping , 2018, 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[13]  Ben Glocker,et al.  Attention Gated Networks: Learning to Leverage Salient Regions in Medical Images , 2018, Medical Image Anal..

[14]  Kwang In Kim,et al.  Unsupervised Attention-guided Image to Image Translation , 2018, NeurIPS.

[15]  Ke Sheng,et al.  Technical Note: Iterative megavoltage CT (MVCT) reconstruction using block‐matching 3D‐transform (BM3D) regularization , 2018, Medical physics.

[16]  Jascha Sohl-Dickstein,et al.  Sensitivity and Generalization in Neural Networks: an Empirical Study , 2018, ICLR.

[17]  Mark Sandler,et al.  CycleGAN, a Master of Steganography , 2017, ArXiv.

[18]  Alexei A. Efros,et al.  Image-to-Image Translation with Conditional Adversarial Networks , 2016, 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[19]  Laurence E. Court,et al.  Learning anatomy changes from patient populations to create artificial CT images for voxel‐level validation of deformable image registration , 2016, Journal of applied clinical medical physics.

[20]  Geoffrey E. Hinton,et al.  Deep Learning , 2015, Nature.

[21]  Andrew Gordon Wilson,et al.  Kernel Interpolation for Scalable Structured Gaussian Processes (KISS-GP) , 2015, ICML.

[22]  Yoshua Bengio,et al.  Generative Adversarial Nets , 2014, NIPS.

[23]  Shuiping Gou,et al.  Denoised and texture enhanced MVCT to improve soft tissue conspicuity. , 2014, Medical physics.

[24]  Emmanuelle Gouillart,et al.  scikit-image: image processing in Python , 2014, PeerJ.

[25]  B G Fallone,et al.  Evaluation of normalized metal artifact reduction (NMAR) in kVCT using MVCT prior images for radiotherapy treatment planning. , 2013, Medical physics.

[26]  Daniel A Low,et al.  Megavoltage CT imaging quality improvement on TomoTherapy via tensor framelet. , 2013, Medical physics.

[27]  Geoffrey E. Hinton,et al.  ImageNet classification with deep convolutional neural networks , 2012, Commun. ACM.

[28]  B. Kavanagh,et al.  High-dose MVCT image guidance for stereotactic body radiation therapy. , 2012, Medical physics.

[29]  Moyed Miften,et al.  Monitoring tumor motion with on-line mega-voltage cone-beam computed tomography imaging in a cine mode , 2008, Physics in medicine and biology.

[30]  H. Zaidi,et al.  Quantitative Analysis in Nuclear Medicine Imaging , 2007, Journal of Nuclear Medicine.

[31]  K. Athreya,et al.  Measure Theory and Probability Theory , 2006 .

[32]  Eero P. Simoncelli,et al.  Image quality assessment: from error visibility to structural similarity , 2004, IEEE Transactions on Image Processing.

[33]  T. Mackie,et al.  Megavoltage CT on a tomotherapy system. , 1999, Physics in medicine and biology.

[34]  Yannis Dimopoulos,et al.  Use of some sensitivity criteria for choosing networks with good generalization ability , 1995, Neural Processing Letters.

[35]  Bernhard Schölkopf,et al.  Kernel methods in medical imaging , 2015 .

[36]  Y. Nesterov.  Gradient methods for minimizing composite functions , 2013, Math. Program..

[37]  Liang Xiao-yang,et al.  CT angiography and CT perfusion imaging guidance in travenous and arterious thrombolytic therapy for 25 cases of stroke progressive , 2010 .

[38]  P. Kupelian,et al.  Megavoltage Computed Tomography Imaging , 2008 .

[39]  Brian F. Hutton,et al.  Iterative Reconstruction Methods , 2006 .

[40]  M. Lev,et al.  17 – CT Angiography and CT Perfusion Imaging , 2002 .

[41]  S. Lang.  Real and Functional Analysis , 1983 .

[42]  A. Rose,et al.  Television Pickup Tubes and the Problem of Vision , 1948 .

[43]  Karl Pearson,et al.  Mathematical Contributions to the Theory of Evolution. XIX. Second Supplement to a Memoir on Skew Variation , 1901 .