Gaussian diffusion sinogram inpainting for X-ray CT metal artifact reduction

Peng, Chengtao; Qiu, Bensheng; Li, Ming; Guan, Yihui; Zhang, Cheng; Wu, Zhongyi; Zheng, Jian

doi:10.1186/s12938-016-0292-9

Cited by 96 publications

(56 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…By interfacing COMSOL with MATLAB (Math Works, USA, http://www.mathwork.com), the calculation process can be completed by using MATLAB, the steps were as follows:

The point coordinates set of each tissue layer were obtained. Since there are critical surfaces between tissue layers, we stipulated that the point on critical surface is assigned into the inner tissue, which is used for obtaining the area of tissues, S bone , S muscle , S fat , S skin .

The FEM was used to calculate the electric field intensity component along the directions of x , y and z , that is E x , E y , and E z , the σ x ( i , j , k ), σ y ( i , j , k ) and σ z ( i , j , k ) can be derived from Gabriel [20], the ρ ( i , j , k ) is equivalent density of whole body in this work [26], whose value is 1000 kg/m 3 .These parameters were substituted into Eq. (4), and the maximum SAR is obtained and then localized SAR can be calculated as the summation of maximum SAR in 10 g continuous tissue.…”

Section: Resultsmentioning

confidence: 99%

“…The FEM was used to calculate the electric field intensity component along the directions of x , y and z , that is E x , E y , and E z , the σ x ( i , j , k ), σ y ( i , j , k ) and σ z ( i , j , k ) can be derived from Gabriel [20], the ρ ( i , j , k ) is equivalent density of whole body in this work [26], whose value is 1000 kg/m 3 .…”

Section: Resultsmentioning

confidence: 99%

“…According to Dirichlet boundary conditions, the input electrical signal at the surface transmitter electrodes is [20] where V 0 is the amplitude of voltage applied to the human arm model by the transmitter electrodes.…”

Section: Methodsmentioning

confidence: 99%

See 2 more Smart Citations

Electrical exposure analysis of galvanic-coupled intra-body communication based on the empirical arm models

et al. 2018

View full text Add to dashboard Cite

BackgroundIntra-body communication (IBC) is one of the highlights in studies of body area networks. The existing IBC studies mainly focus on human channel characteristics of the physical layer, transceiver design for the application, and the protocol design for the networks. However, there are few safety analysis studies of the IBC electrical signals, especially for the galvanic-coupled type. Besides, the human channel model used in most of the studies is just a multi-layer homocentric cylinder model, which cannot accurately approximate the real human tissue layer.MethodsIn this paper, the empirical arm models were established based on the geometrical information of six subjects. The thickness of each tissue layer and the anisotropy of muscle were also taken into account. Considering the International Commission on Non-Ionizing Radiation Protection (ICNIRP) guidelines, the restrictions taken as the evaluation criteria were the electric field intensity lower than 1.35 × 104 f V/m and the specific absorption rate (SAR) lower than 4 W/kg. The physiological electrode LT-1 was adopted in experiments whose size was 4 × 4 cm and the distance between each center of adjoining electrodes was 6 cm. The electric field intensity and localized SAR were all computed by the finite element method (FEM). The electric field intensity was set as average value of all tissues, while SAR was averaged over 10 g contiguous tissue. The computed data were compared with the 2010 ICNIRP guidelines restrictions in order to address the exposure restrictions of galvanic-coupled IBC electrical signals injected into the body with different amplitudes and frequencies.ResultsThe input alternating signal was 1 mA current or 1 V voltage with the frequency range from 10 kHz to 1 MHz. When the subject was stimulated by a 1 mA alternating current, the average electric field intensity of all subjects exceeded restrictions when the frequency was lower than 20 kHz. The maximum difference among six subjects was 1.06 V/m at 10 kHz, and the minimum difference was 0.025 V/m at 400 kHz. While the excitation signal was a 1 V alternating voltage, the electric field intensity fell within the exposure restrictions gradually as the frequency increased beyond 50 kHz. The maximum difference among the six subjects was 2.55 V/m at 20 kHz, and the minimum difference was 0.54 V/m at 1 MHz. In addition, differences between the maximum and the minimum values at each frequency also decreased gradually with the frequency increased in both situations of alternating current and voltage. When SAR was introduced as the criteria, none of the subjects exceeded the restrictions with current injected. However, subjects 2, 4, and 6 did not satisfy the restrictions with voltage applied when the signal amplitude was ≥ 3, 6, and 10 V, respectively. The SAR differences for subjects with different frequencies were 0.062–1.3 W/kg of current input, and 0.648–6.096 W/kg of voltage input.ConclusionBased on the empirical arm models established in this paper, we came to conclusion that the frequ...

show abstract

“…By interfacing COMSOL with MATLAB (Math Works, USA, http://www.mathwork.com), the calculation process can be completed by using MATLAB, the steps were as follows:

Section: Resultsmentioning

confidence: 99%

Section: Resultsmentioning

confidence: 99%

See 1 more Smart Citation

Electrical exposure analysis of galvanic-coupled intra-body communication based on the empirical arm models

et al. 2018

View full text Add to dashboard Cite

show abstract

“…Metal-corrupted CT sinograms are synthesized using the dynamic wavelet thresholding method for x-ray CT metal artifact reduction by Peng et al 19 and the dual domain network method for CT metal artifact reduction by Lin et al 31 In addition, to test the clinical efficacy of the proposed method, we evaluate it on a real clinical data (a CT image that contains real metal artifacts), which is shared by the authors of another MAR study. 9 The size of the clinical CT image is 512 9 512.…”

Section: Datasetsmentioning

confidence: 99%

An irregular metal trace inpainting network for x‐ray CT metal artifact reduction

Peng

et al. 2020

Medical Physics

Self Cite

View full text Add to dashboard Cite

Purpose: Metal implants in the patient's body can generate severe metal artifacts in x-ray computed tomography (CT) images. These artifacts may cover the tissues around the metal implants in CT images and even corrupt the tissue regions, thus affecting disease diagnosis using these images. Previous deep learning metal trace inpainting methods used both valid pixels of uncorrupted areas and invalid pixels of corrupted areas to patch metal trace (i.e., the holes of removed metal-corrupted regions). Such methods cannot recover fine details well and often suffer information mismatch due to interference of invalid pixels, thus incurring considerable secondary artifacts. In this paper, we develop a new irregular metal trace inpainting network for reducing metal artifacts. Methods: We develop a new deep learning network to patch irregular metal trace in metal-corrupted sinograms to reduce metal artifacts for isometric fan-beam CT. Our new method patches irregular metal trace in CT sinograms using only valid pixels, avoiding interference from invalid pixels. Furthermore, to enable the inpainting network to recover as many details as possible, we design an auxiliary inpainting network to suppress the probable secondary artifacts in CT images to assist fine detail restoration. The image produced by the auxiliary network is then projected onto a sinogram via a forward projection (FP) algorithm and is fused with the sinogram predicted by the inpainting network in order to predict the final recovered sinogram. Our entire network is trained end-to-end to extract cross-domain information between the sinogram domain and CT image domain. Results: We compare our proposed method with two traditional and four deep learning-based metal trace inpainting methods, and with an iterative reconstruction method on four datasets: dental fillings (panoramic and local perspectives), hip prostheses, and spine fixations. We use both quantitative and qualitative indices to evaluate our method, and the analyses suggest that our method reduces the most metal artifacts and produces the best quality CT images. Additionally, our proposed method takes 0.1512 s on average to process a CT slice, which meets the clinical requirement. Conclusions: This paper proposes a new deep learning network to patch irregular metal trace in corrupted sinograms to reduce metal artifacts. Our method restores more fine details in irregular metal trace and has a superior capability on metal artifact reduction compared with state-of-the-art methods.

show abstract

“…Image artifacts caused by the presence of metals sometimes affect the interpretation of computed tomography (CT) images [1,2] and the planning of radiation therapy [3]. Several metal artifact reduction (MAR) techniques have been developed to minimize metal artifacts in CT [4][5][6][7][8]. Four commercially available MAR techniques are single-energy MAR (SEMAR); gemstone spectral imaging (GSI) dual-energy CT (DECT) with MAR software (MARS); iterative MAR (IMAR); and orthopedic MAR (OMAR).…”

Section: Introductionmentioning

confidence: 99%

Quantitative assessment of three vendor’s metal artifact reduction techniques for CT imaging using a customized phantom

Chou

Ying

et al. 2019

Computer Assisted Surgery

View full text Add to dashboard Cite

A metal implant was placed in an acrylic phantom to enable quantitative analysis of the metal artifact reduction techniques used in computed tomography (CT) scanners from three manufacturers. Two titanium rods were placed in a groove in a cylindrical phantom made by acrylic, after which the groove was filled with water. The phantom was scanned using three CT scanners (Toshiba, GE, Siemens) under the abdomen CT setting. CT number accuracy, contrast-tonoise ratio, area of the metal rods in the images, and fraction of affected pixel area of water were measured using ImageJ. Different iterative reconstruction, dual energy, and metal artifact reduction techniques were compared within three vendors. The highest contrast-to-noise ratio of three scanners were 85.7 ± 8.4 (Toshiba), 85.9 ± 11.7 (GE), and 55.0 ± 14.8 (Siemens); and the most correct results of metal area were 157.1 ± 1.4 mm 2 (Toshiba), 155.0 ± 1.0 (GE), and 170.6 ± 5.3 (Siemens). The fraction of affected pixel area obtained using single-energy metal artifact reduction of Toshiba scanner was 2.2% ± 0.7%, which is more favorable than 4.1% ± 0.7% obtained using metal artifact reduction software of GE scanner (p ¼ 0.002). Among all quantitative results, the estimations with fraction of affected pixel areas matched the effect of metal artifact reduction in the actual images. Therefore, the single-energy metal artifact reduction technique of Toshiba scanner had a desirable effect. The metal artifact reduction software of GE scanner effectively reduced the effect of metal artifacts; however, it underestimated the size of the metal rods. The monoenergetic and dual energy composition techniques of Siemens scanner could not effectively reduce metal artifacts.

show abstract

Gaussian diffusion sinogram inpainting for X-ray CT metal artifact reduction

Cited by 96 publications

References 22 publications

Electrical exposure analysis of galvanic-coupled intra-body communication based on the empirical arm models

Electrical exposure analysis of galvanic-coupled intra-body communication based on the empirical arm models

An irregular metal trace inpainting network for x‐ray CT metal artifact reduction

Quantitative assessment of three vendor’s metal artifact reduction techniques for CT imaging using a customized phantom

Contact Info

Product

Resources

About