Chest tomosynthesis deblurring using CNN with deconvolution layer for vertebrae segmentation.
CNN
Deblur
Tomosynthesis
Journal
Medical physics
ISSN: 2473-4209
Titre abrégé: Med Phys
Pays: United States
ID NLM: 0425746
Informations de publication
Date de publication:
Dec 2023
Dec 2023
Historique:
revised:
13
04
2023
received:
17
11
2022
accepted:
06
06
2023
medline:
6
12
2023
pubmed:
4
7
2023
entrez:
4
7
2023
Statut:
ppublish
Résumé
Limited scan angles cause severe distortions and artifacts in reconstructed tomosynthesis images when the Feldkamp-Davis-Kress (FDK) algorithm is used for the purpose, which degrades clinical diagnostic performance. These blurring artifacts are fatal in chest tomosynthesis images because precise vertebrae segmentation is crucial for various diagnostic analyses, such as early diagnosis, surgical planning, and injury detection. Moreover, because most spinal pathologies are related to vertebral conditions, the development of methods for accurate and objective vertebrae segmentation in medical images is an important and challenging research area. The existing point-spread-function-(PSF)-based deblurring methods use the same PSF in all sub-volumes without considering the spatially varying property of tomosynthesis images. This increases the PSF estimation error, thus further degrading the deblurring performance. However, the proposed method estimates the PSF more accurately by using sub-CNNs that contain a deconvolution layer for each sub-system, which improves the deblurring performance. To minimize the effect of the spatially varying property, the proposed deblurring network architecture comprises four modules: (1) block division module, (2) partial PSF module, (3) deblurring block module, and (4) assembling block module. We compared the proposed DL-based method with the FDK algorithm, total-variation iterative reconstruction with GP-BB (TV-IR), 3D U-Net, FBPConvNet, and two-phase deblurring method. To investigate the deblurring performance of the proposed method, we evaluated its vertebrae segmentation performance by comparing the pixel accuracy (PA), intersection-over-union (IoU), and F-score values of reference images to those of the deblurred images. Also, pixel-based evaluations of the reference and deblurred images were performed by comparing their root mean squared error (RMSE) and visual information fidelity (VIF) values. In addition, 2D analysis of the deblurred images were performed by artifact spread function (ASF) and full width half maximum (FWHM) of the ASF curve. The proposed method was able to recover the original structure significantly, thereby further improving the image quality. The proposed method yielded the best deblurring performance in terms of vertebrae segmentation and similarity. The IoU, F-score, and VIF values of the chest tomosynthesis images reconstructed using the proposed SV method were 53.5%, 28.7%, and 63.2% higher, respectively, than those of the images reconstructed using the FDK method, and the RMSE value was 80.3% lower. These quantitative results indicate that the proposed method can effectively restore both the vertebrae and the surrounding soft tissue. We proposed a chest tomosynthesis deblurring technique for vertebrae segmentation by considering the spatially varying property of tomosynthesis systems. The results of quantitative evaluations indicated that the vertebrae segmentation performance of the proposed method was better than those of the existing deblurring methods.
Sections du résumé
BACKGROUND
BACKGROUND
Limited scan angles cause severe distortions and artifacts in reconstructed tomosynthesis images when the Feldkamp-Davis-Kress (FDK) algorithm is used for the purpose, which degrades clinical diagnostic performance. These blurring artifacts are fatal in chest tomosynthesis images because precise vertebrae segmentation is crucial for various diagnostic analyses, such as early diagnosis, surgical planning, and injury detection. Moreover, because most spinal pathologies are related to vertebral conditions, the development of methods for accurate and objective vertebrae segmentation in medical images is an important and challenging research area.
PURPOSE
OBJECTIVE
The existing point-spread-function-(PSF)-based deblurring methods use the same PSF in all sub-volumes without considering the spatially varying property of tomosynthesis images. This increases the PSF estimation error, thus further degrading the deblurring performance. However, the proposed method estimates the PSF more accurately by using sub-CNNs that contain a deconvolution layer for each sub-system, which improves the deblurring performance.
METHODS
METHODS
To minimize the effect of the spatially varying property, the proposed deblurring network architecture comprises four modules: (1) block division module, (2) partial PSF module, (3) deblurring block module, and (4) assembling block module. We compared the proposed DL-based method with the FDK algorithm, total-variation iterative reconstruction with GP-BB (TV-IR), 3D U-Net, FBPConvNet, and two-phase deblurring method. To investigate the deblurring performance of the proposed method, we evaluated its vertebrae segmentation performance by comparing the pixel accuracy (PA), intersection-over-union (IoU), and F-score values of reference images to those of the deblurred images. Also, pixel-based evaluations of the reference and deblurred images were performed by comparing their root mean squared error (RMSE) and visual information fidelity (VIF) values. In addition, 2D analysis of the deblurred images were performed by artifact spread function (ASF) and full width half maximum (FWHM) of the ASF curve.
RESULTS
RESULTS
The proposed method was able to recover the original structure significantly, thereby further improving the image quality. The proposed method yielded the best deblurring performance in terms of vertebrae segmentation and similarity. The IoU, F-score, and VIF values of the chest tomosynthesis images reconstructed using the proposed SV method were 53.5%, 28.7%, and 63.2% higher, respectively, than those of the images reconstructed using the FDK method, and the RMSE value was 80.3% lower. These quantitative results indicate that the proposed method can effectively restore both the vertebrae and the surrounding soft tissue.
CONCLUSIONS
CONCLUSIONS
We proposed a chest tomosynthesis deblurring technique for vertebrae segmentation by considering the spatially varying property of tomosynthesis systems. The results of quantitative evaluations indicated that the vertebrae segmentation performance of the proposed method was better than those of the existing deblurring methods.
Types de publication
Journal Article
Langues
eng
Sous-ensembles de citation
IM
Pagination
7714-7730Subventions
Organisme : Ministry of Science and ICT, South Korea
ID : RS-2022-00144336
Organisme : Ministry of Science and ICT, South Korea
ID : RS-2023-00219019
Organisme : Ministry of Science and ICT, South Korea
ID : RS-2023-00240135
Informations de copyright
© 2023 American Association of Physicists in Medicine.
Références
Dobbins III JT, McAdams HP. Chest tomosynthesis: technical principles and clinical update. Eur J Radiol. 2009;72:244-251.
Duryea J, Dobbins III J, Lynch J. Digital tomosynthesis of hand joints for arthritis assessment. Med Phys. 2003;30:325-333.
Al-Mokhtar N, Shah J, Marson B, Evans S, Nye K. Initial clinical experience of the use of digital tomosynthesis in the assessment of suspected fracture neck of femur in the elderly. Eur J Orthop Surg Traumatol. 2015;25:941-947.
Cho M, Kim H, Youn H, Kim S. A feasibility study of digital tomosynthesis for volumetric dental imaging. J Instrum. 2012;7:P03007.
Gennaro G, Toledano A, di Maggio C, et al. Digital breast tomosynthesis versus digital mammography: a clinical performance study. Eur Radiol. 2010;20:1545-1553.
Sechopoulos I. A review of breast tomosynthesis. Part I. The image acquisition process. Med Phys. 2013;40:014301.
Reiser I, Nishikawa R, Giger M, et al. Computerized mass detection for digital breast tomosynthesis directly from the projection images. Med Phys. 2006;33:482-491.
Zhang Y, Chan H-P, Sahiner B, et al. A comparative study of limited-angle cone-beam reconstruction methods for breast tomosynthesis. Med Phys. 2006;33:3781-3795.
Gong X, Glick SJ, Liu B, Vedula AA, Thacker S. A computer simulation study comparing lesion detection accuracy with digital mammography, breast tomosynthesis, and cone-beam CT breast imaging. Med Phys. 2006;33:1041-1052.
Korez R, Ibragimov B, Likar B, Pernuš F, Vrtovec T. A framework for automated spine and vertebrae interpolation-based detection and model-based segmentation. IEEE Trans Med Imaging. 2015;34:1649-1662.
Al Arif S, Knapp K, Slabaugh G. Shape-aware deep convolutional neural network for vertebrae segmentation. In: International Workshop on Computational Methods and Clinical Applications in Musculoskeletal Imaging. Springer; 2017:12-24.
Hammernik K, Ebner T, Stern D, Urschler M, Pock T. Vertebrae segmentation in 3D CT images based on a variational framework. In: Recent advances in computational methods and clinical applications for spine imaging. Springer; 2015:227-233.
Al Arif SMR, Knapp K, Slabaugh G. Fully automatic cervical vertebrae segmentation framework for X-ray images. Comput Methods Programs Biomed. 2018;157:95-111.
Rose SD, Sidky EY, Reiser IS, Pan X. Filtered back-projection for digital breast tomosynthesis with 2D filtering. In: Medical Imaging 2019: Physics of Medical Imaging. Vol. 10948. SPIE; 2019:1193-1196.
Park JC, Song B, Kim JS, et al. Fast compressed sensing-based CBCT reconstruction using Barzilai-Borwein formulation for application to on-line IGRT. Med Phys. 2012;39:1207-1217.
Xu J, Zhao Y, Li H, Zhang P. An image reconstruction model regularized by edge-preserving diffusion and smoothing for limited-angle computed tomography. Inverse Prob. 2019;35:085004.
Hu Y-H, Zhao B, Zhao W. Image artifacts in digital breast tomosynthesis: investigation of the effects of system geometry and reconstruction parameters using a linear system approach. Med Phys. 2008;35:5242-5252.
Mota AM, Clarkson MJ, Almeida P, Matela N. An enhanced visualization of DBT imaging using blind deconvolution and total variation minimization regularization. IEEE Trans Med Imaging. 2020;39:4094-4101.
Chen H, Zhang Y, Kalra MK, et al. Low-dose CT with a residual encoder-decoder convolutional neural network. IEEE Trans Med Imaging. 2017;36:2524-2535.
Zhang Y, Yu H. Convolutional neural network based metal artifact reduction in x-ray computed tomography. IEEE Trans Med Imaging. 2018;37:1370-1381.
Ko Y, Moon S, Baek J, Shim H. Rigid and non-rigid motion artifact reduction in X-ray CT using attention module. Med Image Anal. 2021;67:101883.
Ye DH, Buzzard GT, Ruby M, Bouman CA. Deep back projection for sparse-view CT reconstruction. In: 2018 IEEE Global Conference on Signal and Information Processing (GlobalSIP). IEEE; 2018:1-5.
Choi Y, Han M, Jang H, Shim H, Baek J. Two-phase learning-based 3D deblurring method for digital breast tomosynthesis images. PloS one. 2022;17:e0262736.
Jiang Z, Yin F-F, Ge Y, Ren L. Enhancing digital tomosynthesis (DTS) for lung radiotherapy guidance using patient-specific deep learning model. Phys Med Biol. 2021;66:035009.
He J, Yang Y, Wang Y, et al. Optimizing a parameterized plug-and-play ADMM for iterative low-dose CT reconstruction. IEEE Trans Med Imaging. 2018;38:371-382.
Cheng J, Wang H, Zhu Y, et al. Model-based deep medical imaging: the roadmap of generalizing iterative reconstruction model using deep learning. arXiv preprint arXiv:1906.08143 (2019).
Su T, Deng X, Yang J, et al. DIR-DBTnet: Deep iterative reconstruction network for three-dimensional digital breast tomosynthesis imaging. Med Phys. 2021;48:2289-2300.
Li B, Avinash GB, Eberhard JW, Claus BE. Optimization of slice sensitivity profile for radiographic tomosynthesis. Med Phys. 2007;34:2907-2916.
Zeiler MD, Taylor GW, Fergus R. Adaptive deconvolutional networks for mid and high level feature learning. In: 2011 international conference on computer vision. IEEE; 2011:2018-2025.
Çiçek Ö, Abdulkadir A, Lienkamp SS, Brox T, Ronneberger O. 3D U-Net: learning dense volumetric segmentation from sparse annotation. In: International conference on medical image computing and computer-assisted intervention. Springer; 2016:424-432.
Jin KH, McCann MT, Froustey E, Unser M. Deep convolutional neural network for inverse problems in imaging. IEEE Trans Image Process. 2017;26:4509-4522.
Jaccard P. The distribution of the flora in the alpine zone. New Phytol. 1912;11:37-50.
Sasaki Y, et al. The truth of the F-measure. Teach Tutor Mater. 2007;1:1-5.
Sheikh HR, Bovik AC. A visual information fidelity approach to video quality assessment. In: The First International Workshop on Video Processing and Quality Metrics for Consumer Electronics. Vol 7. sn; 2005:2117-2128.
Siddon RL. Fast calculation of the exact radiological path for a three-dimensional CT array. Med Phys. 1985;12:252-255.
Dobbins III JT, McAdams HP, Song J-W, et al. Digital tomosynthesis of the chest for lung nodule detection: Interim sensitivity results from an ongoing NIH-sponsored trial. Med Phys. 2008;35:2554-2557.
Zeiler MD, Krishnan D, Taylor GW, Fergus R. Deconvolutional networks. In: 2010 IEEE Computer Society Conference on computer vision and pattern recognition. IEEE; 2010:2528-2535.
Kingma DP, Ba J. Adam: A method for stochastic optimization. arXiv preprint arXiv:1412.6980 (2014).
Simonyan K, Zisserman A. Very deep convolutional networks for large-scale image recognition. arXiv preprint arXiv:1409.1556 (2014).
Chan TF, Vese LA. Active contours without edges. IEEE Trans Image Process. 2001;10:266-277.
Kim Y-D, Park E, Yoo S, Choi T, Yang L, Shin D. Compression of deep convolutional neural networks for fast and low power mobile applications. arXiv preprint arXiv:1511.06530 (2015).
Ding H, Chen K, Yuan Y, et al. A compact CNN-DBLSTM based character model for offline handwriting recognition with Tucker decomposition. In: 2017 14th IAPR International Conference on Document Analysis and Recognition (ICDAR). Vol 1. IEEE; 2017:507-512.
Li Z, Ren A, Li J, et al. Structural design optimization for deep convolutional neural networks using stochastic computing. In: Design, Automation & Test in Europe Conference & Exhibition (DATE), 2017. IEEE; 2017:250-253.
Ren A, Li Z, Ding C, et al. Sc-dcnn: Highly-scalable deep convolutional neural network using stochastic computing. ACM SIGPLAN Notices 2017;52:405-418.
Xiao Q, Liang Y, Lu L, Yan S, Tai Y-W. Exploring heterogeneous algorithms for accelerating deep convolutional neural networks on FPGAs. In: Proceedings of the 54th Annual Design Automation Conference 2017. Association for Computing Machinery; 2017:1-6.
Kalender WA, Hebel R, Ebersberger J. Reduction of CT artifacts caused by metallic implants. Radiology 1987;164:576-577.
Meyer E, Raupach R, Lell M, Schmidt B, Kachelrieß M. Normalized metal artifact reduction (NMAR) in computed tomography. Med Phys. 2010;37:5482-5493.
Lee C, Song H-d, Baek J. 3D MTF estimation using sphere phantoms for cone-beam computed tomography systems. Med Phys. 2020;47:2838-2851.
Chen Z, Ning R. Three-dimensional point spread function measurement of cone-beam computed tomography system by iterative edge-blurring algorithm. Phys Med Biol. 2004;49:1865.
Thornton MM, Flynn MJ. Measurement of the spatial resolution of a clinical volumetric computed tomography scanner using a sphere phantom. In: Medical Imaging 2006: Physics of Medical Imaging. Vol 6142. SPIE; 2006:707-716.
Samei E, Murphy S, Richard S. Assessment of multi-directional MTF for breast tomosynthesis. Phys Med Biol. 2013;58:1649.
Steiding C, Kolditz D, Kalender WA. A quality assurance framework for the fully automated and objective evaluation of image quality in cone-beam computed tomography. Med Phys. 2014;41:031901.
Stayman JW, Capostagno S, Gang GJ, Siewerdsen JH. Task-driven source-detector trajectories in cone-beam computed tomography: I. Theory and methods. J Med Imaging. 2019;6:025002.
Lee C, Baek J. A new method to measure directional modulation transfer function using sphere phantoms in a cone beam computed tomography system. IEEE Trans Med Imaging. 2014;34:902-910.
Robert N, Mainprize JG, Whyne C. Determination of 3D PSFs from computed tomography reconstructed x-ray images of spherical objects and the effects of sphere radii. Med Phys. 2019;46:4792-4802.
Gulrajani I, Ahmed F, Arjovsky M, Dumoulin V, Courville AC. Improved training of wasserstein gans. Adv Neural Inf Process Syst. 2017;30.
Yang Q, Yan P, Zhang Y, et al. Low-dose CT image denoising using a generative adversarial network with Wasserstein distance and perceptual loss. IEEE Trans Med Imaging. 2018;37:1348-1357.
Han M, Shim H, Baek J. Low-dose CT denoising via convolutional neural network with an observer loss function. Med Phys. 2021;48:5727-5742.
Zhu J-Y, Park T, Isola P, Efros AA. Unpaired image-to-image translation using cycle-consistent adversarial networks. In: Proceedings of the IEEE international conference on computer vision. IEEE; 2017:2223-2232.
Mildenhall B, Srinivasan PP, Tancik M, Barron JT, Ramamoorthi R, Ng R. Nerf: Representing scenes as neural radiance fields for view synthesis. In: European conference on computer vision. Springer; 2020:405-421.
Sim B, Oh G, Kim J, Jung C, Ye JC. Optimal transport driven CycleGAN for unsupervised learning in inverse problems. SIAM J Imaging Sci. 2020;13:2281-2306.