X-ray Computed Laminography (CL) is a popular industrial tool for non-destructive visualization of flat objects. However, high-quality CL imaging requires a large number of projections, resulting in a long imaging time. Reducing the number of projections allows acceleration of the imaging process, but decreases the quality of reconstructed images.
Objective
Our objective is to build a deep learning network for sparse-view CL reconstruction.
Methods
Considering complementarities of feature extraction in different domains, we design an encoder-decoder network that enables to compensate the missing information during spatial domain feature extraction in wavelet domain. Also, a detail-enhanced module is developed to highlight details. Additionally, Swin Transformer and convolution operators are combined to better capture features.
Results
A total of 3200 pairs of 16-view and 1024-view CL images (2880 pairs for training, 160 pairs for validation, and 160 pairs for testing) of solder joints have been employed to investigate the performance of the proposed network. It is observed that the proposed network obtains the highest image quality with PSNR and SSIM of 37.875 ± 0.908 dB, 0.992 ± 0.002, respectively. Also, it achieves competitive results on the AAPM dataset.
Conclusions
This study demonstrates the effectiveness and generalization of the proposed network for sparse-view CL reconstruction.
MaislMPorschFSchorrC. Computed laminography for x-ray inspection of lightweight constructions. In: 2nd international symposium on NDT in aerospace 2010, Vol. 3. 2010, pp.1–7.
2.
WangZLiuYWangS, et al.Analytical reconstructions of full-scan multiple source-translation computed tomography under large field of views. J X-Ray Sci Technol2023; 31: 1245–1262.
3.
O’BrienNSBoardmanRPSinclairI, et al.Recent advances in X-ray cone-beam computed laminography. J X-ray Sci Technol2016; 24: 691–707.
4.
JiaTShiLWeiC, et al.Correction of motion artifact in CL based on MAFusNet. J X-Ray Sci Technol2023; 31: 393–407.
5.
SunYHanYTanS, et al.Geometric parameters sensitivity evaluation based on projection trajectories for X-ray cone-beam computed laminography. J X-Ray Sci Technol2023; 31: 423–434.
6.
QueJMCaoDQZhaoW, et al.Computed laminography and reconstruction algorithm. Chin Phys C2012; 36: 77.
7.
YangMWangGLiuY. New reconstruction method for x-ray testing of multilayer printed circuit board. Opt Eng2010; 49: 056501-056501-6.
8.
SunLZhouGQinZ, et al.A reconstruction method for cone-beam computed laminography based on projection transformation. Meas Sci Technol2021; 32: 045403.
9.
ParkMChoGParkJ, et al.Noncircular scanning trajectory for sparse-view computed laminography. In: 6th conference on industrial computed tomography. 2016.
10.
ParkMLeeMKimK, et al.Scanning parameter optimization for inspection of wedling defects in X-ray linear laminography. In: International conference on industrial computed tomography. 2017.
11.
AbbasSParkMMinJ, et al.Sparse-view computed laminography with a spherical sinusoidal scan for nondestructive testing. Opt Exp2014: 17745–17755. doi: https://doi.org/10.1364/OE.22.017745.
12.
AbbasSParkMChoS. Fast and low-dose computed laminography using compressive sensing based technique. AIP Conf Proc2015: 586–590. doi: https://doi.org/10.1063/1.4914657.
13.
BatenburgKJPalenstijnWJSijbersJ. 3D Imaging of semiconductor components by discrete laminography. AIP Conf Proc2014; 1601: 168–179.
14.
BatenburgKJSijbersJ. DART: a practical reconstruction algorithm for discrete tomography. IEEE Trans Image Process2011; 20: 2542–2553.
15.
ChanYLiuXWangT, et al.An attention-based deep convolutional neural network for ultra-sparse-view CT reconstruction. Comput Biol Med2023; 161: 106888.
16.
LeeHLeeJChoS. View-interpolation of sparsely sampled sinogram using convolutional neural network. SPIE2017: 617–624. doi: https://doi.org/10.1117/12.2254244.
17.
LeeHLeeJKimH, et al.Deep-neural-network-based sinogram synthesis for sparse-view CT image reconstruction. IEEE Trans Radiat Plasma Med Sci2018; 3: 109–119.
18.
XieSZhengXChenY, et al.Artifact removal using improved GoogLeNet for sparse-view CT reconstruction. Sci Rep2018; 8: 1–9.
19.
ZhangZLiangXDongX, et al.A sparse-view CT reconstruction method based on combination of DenseNet and deconvolution. IEEE Trans Med Imag2018; 37: 1407–1417.
20.
LinWALiaoHPengC, et al.Dudonet: dual domain network for ct metal artifact reduction. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. 2019, pp.10512–10521. doi: https://doi.org/10.1109/CVPR.2019.01076.
21.
TangHJiangSLinY, et al.An improved dual-domain network for metal artifact reduction in CT images using aggregated contextual transformations. Phys Med Biol2023; 68: 175021.
22.
PereiraLFADe BeenhouwerJKastnerJ, et al.Extreme sparse X-ray computed laminography via convolutional neural networks. In: 2020 IEEE 32nd international conference on tools with artificial intelligence (ICTAI), Vol. 2020. 2020, pp.612–616. doi: https://doi.org/10.1109/ICTAI50040.2020.00099.
23.
KangIJiangYHollerM, et al.Accelerated deep self-supervised ptycho-laminography for three-dimensional nanoscale imaging of integrated circuits. Optica2023; 10: 1000–1008.
24.
LiuZLinYCaoY, et al.Swin transformer: hierarchical vision transformer using shifted windows. In: Proceedings of the IEEE/CVF international conference on computer vision, Vol. 2021. 2021, pp.10012–10022. doi: https://doi.org/10.1109/ICCV48922.2021.00986.
ShiLQuG. A preconditioned landweber iteration scheme for the limited-angle image reconstruction. J X-Ray Sci Technol2021; 29: 1045–1063.
27.
CaoHWangYChenJ, et al.Swin-unet: unet-like pure transformer for medical image segmentation. In: European conference on computer vision. Cham: Springer Nature Switzerland, 2022, pp.205–218. doi: https://doi.org/10.1007/978-3-031-25066-8_9.
28.
SobelI.An Isotropic 3× 3 Image Gradient Operator, Presentation at Stanford AI Project. 2014.
29.
MainiRSohalJS. Performance evaluation of prewitt edge detector for noisy images. GVIP J2006; 6: 39–46.
30.
KnausCZwickerM.Dual-domain image denoising. In: 2013 IEEE international conference on image processing, Vol. 2013. 2013, pp.440–444. doi: https://doi.org/10.1109/ICIP.2013.6738091.
31.
KanagarajHMuneeswaranV.Image compression using HAAR discrete wavelet transform. In: 2020 5th international conference on devices, circuits and systems (ICDCS), Vol. 2020. IEEE, 2020, pp.271–274. doi: https://doi.org/10.1109/ICDCS48716.2020.243596.
32.
JinKHMcCannMTFrousteyE, et al.Deep convolutional neural network for inverse problems in imaging. IEEE Trans Image Process2017; 26: 4509–4522.
33.
LeeMKimHKimHJ. Sparse-view CT reconstruction based on multi-level wavelet convolution neural network. Phys Med2020; 80: 352–362.
34.
McColloughC. TU-FG-207a-04: overview of the low dose CT grand challenge. Med Phys2016; 43: 3759–3760. doi: https://doi.org/10.1118/1.4957556.
35.
ZhangYHuDHaoS, et al.DREAM-Net: deep residual error iterative minimization network for sparse-view CT reconstruction. IEEE J Biomed Health Inf2022; 27: 480–491.
