Super resolution dual-energy cone-beam CT imaging with dual-layer ﬂat-panel detector Ting Su1Jiongtao Zhu2Xin Zhang1Dong Zeng3Yuhang Tan1Han Cui1Hairong

2025-04-26 0 0 2.04MB 30 页 10玖币

侵权投诉

Super resolution dual-energy cone-beam CT imaging with dual-layer ﬂat-panel

detector

Ting Su,1Jiongtao Zhu,2Xin Zhang,1Dong Zeng,3Yuhang Tan,1Han Cui,1Hairong

Zheng,4Jianhua Ma,3Dong Liang,1, 4 and Yongshuai Ge1, 4, a)

1)Research Center for Medical Artiﬁcial Intelligence, Shenzhen Institutes of

Advanced Technology, Chinese Academy of Sciences, Shenzhen, Guangdong 518055,

China.

2)College of Physics and Optoelectronic Engineering, Ministry of Education and

Guangdong Province, Key Laboratory of Optoelectronic Equipment and Systems,

Shenzhen University, Shenzhen 518060, China.

3)School of Biomedical Engineering, Southern Medical University,

Guangzhou 510515, China.

4)Paul C Lauterbur Research Center for Biomedical Imaging,

Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences,

Shenzhen, Guangdong 518055, China.

(Dated: 18 October 2022)

arXiv:2210.05884v2 [physics.med-ph] 17 Oct 2022

For medical cone-beam computed tomography (CBCT) imaging, the native recep-

tor array of the ﬂat-panel detector (FPD) is usually binned into a reduced matrix

size. By doing so, the signal readout speed can be increased by over 4-9 times at

the expense of sacriﬁcing the spatial resolution by at least 50%-67%. Clearly, such

tradition poses a main bottleneck in generating high spatial resolution and high tem-

poral resolution CBCT images at the same time. In addition, the conventional FPD

is also diﬃcult in generating dual-energy CBCT images. In this paper, we propose

an innovative super resolution dual-energy CBCT imaging method, named as suRi,

based on dual-layer FPD (DL-FPD) to overcome these aforementioned diﬃculties

at once. With suRi, speciﬁcally, an 1D or 2D sub-pixel (half pixel in this study)

shifted binning is applied to replace the conventionally aligned binning to double

the spatial sampling rate during the dual-energy data acquisition. As a result, the

suRi approach provides a new strategy to enable high signal readout speed and high

spatial resolution CBCT imaging with FPD. Moreover, a penalized likelihood mate-

rial decomposition algorithm is developed to directly reconstruct the high resolution

bases from the dual-energy CBCT projections containing spatial sub-pixel shifts. Ex-

periments based on the single-layer FPD and DL-FPD are performed with physical

phantoms and biological specimen to validate this newly developed suRi method.

The synthesized monochromatic CT imaging results demonstrate that suRi can sig-

niﬁcantly improve the spatial image resolution by 46.15%. We believe the developed

suRi method would be capable to greatly enhance the imaging performance of the

DL-FPD based dual-energy CBCT systems in future.

Keywords: System modeling, dual-energy imaging, sub-pixel shift, CT image recon-

struction.

a)Electronic mail: ys.ge@siat.ac.cn.

I. INTRODUCTION

Over the past two decades, X-ray ﬂat-panel detector (FPD) made from a single layer

of CsI:TI scintillator and amorphous silicon (α-Si) based thin ﬁlm transistor (TFT) back-

plate has been widely used in medical imaging applications. Due to its small native pixel

dimension, e.g., 0.07-0.2 mm, FPD shows superior performance in detecting the ultra-ﬁne

details of the anatomical structures such as the micro-calciﬁcation, bone marrow and contrast

enhanced blood vessels in two dimensional digital radiography (DR) imaging. For three and

four dimensional cone-beam computed tomography (CBCT) imaging, ﬂat-panel detector

(FPD) is also irreplaceable in applications such as oral imaging1, image-guided radiation

therapy2and interventional therapy3. Despite of these advancements, however, the dilemma

between the intrinsic pixel-size-deﬁned high spatial resolution and the pixel-number-conﬁned

slow data acquisition speed in FPD impedes its further developments in advanced CBCT

imaging tasks which require both high spatial and high temporal resolution simultaneously.

Take the PaxScan 4030CB FPD (Varex, USA) as an example, the native pixel dimension is

0.194 mm, and the full receptor array consists of 2048 ×1536 detector elements. With the

1×1binning mode (limiting resolution 2.58 lp/mm), this FPD can only acquire 7.5 frames

per second (fps) at most4,5, and at least 40 seconds are needed to complete a CBCT scan.

The excessive CBCT scanning time would cause the motion artifacts and the loss of temporal

variations of the contrast agent. Therefore, the FPD is usually worked at 2×2binning mode

(0.388 mm eﬀective pixel dimension, corresponding to a reduced limiting spatial resolution

of 1.29 lp/mm). By doing so, a quick signal readout speed of 30.0 fps can be achieved to

complete the CBCT data acquisition within 6-10 seconds. Compared to the 1×1binning

mode, the 2×2binning mode saves about 80% CBCT scan time, but at the expense of

losing 50% image spatial resolution4. Besides, the current FPD is also diﬃcult in performing

dual-energy CBCT imaging to generate quantitative material-speciﬁc images for accurate

disease diagnoses6.

As a promising dual-energy CBCT imaging approach, the ﬂat-panel detector made from

dual layers of CsI:TI scintillator and α-Si TFT back-plate (DL-FPD) can acquire the dual-

energy CBCT data simultaneously: the low-energy (LE) X-ray projection is acquired from

the top layer, and the high-energy (HE) X-ray projection is acquired from the bottom

layer. The two CsI:TI scintillator layers may have diﬀerent thicknesses, and additional beam

ﬁltration made of 1.0 mm Copper may also be inserted between the two layers to further

separate the beam spectra. By far, several investigations based on the DL-FPD have been

reported. For example, Shi et al. and Ståhl et al. demonstrated the feasibility of performing

accurate dual-energy CBCT imaging based on the DL-FPD7,8. Wang et al. proposed a

model-based high resolution material decomposition method for the DL-FPD CBCT9. Even

though the DL-FPD helps realizing the dual-energy CBCT imaging, however, the conﬂict

between the high spatial resolution imaging and the high speed imaging in single-layer FPD

(SL-FPD) is still inherited. In other words, the DL-FPD encounters the same trade-oﬀ

between the spatial resolution and the temporal resolution as the SL-FPD.

To overcome this long-standing diﬃculty, we propose an innovative imaging method,

named as suRi, for DL-FPD to realize high temporal and spatial resolution CBCT imaging

at the same time. In particular, the conventionally aligned binning approach is altered into a

sub-pixel (half pixel in this study) shifted binning approach, see the illustrations in Fig. 1for

more details. Depending on the imaging task, such half pixel shift could be one-dimensional

along the horizontal direction, or two-dimensional along the diagonal direction. Take the

simple one-dimensional half pixel shift as an example, the information recorded on the top

and bottom layers are spatially shifted by half-pixel. As shown in Fig. 1, the line integral

of the polychromatic X-ray beam that passes through a certain object position is sampled

by two diﬀerent FPD pixel elements: the top one is shifted by half pixel with regard to the

bottom one. Consequently, the sub-pixel shift used by suRi doubles the spatial sampling

rate assuming parallel incident beam. Therefore, the proposed suRi approach not only

enables fast dual-energy CBCT imaging, but also enables super spatial resolution CBCT

imaging. With such redeﬁned DL-FPD data acquisition scheme, we believe the dual-energy

CBCT imaging with high spatial resolution and high temporal resolution performance can

be achieved. To the best of our knowledge, no such investigations have been reported before.

The major contributions of this work are as follows: (1) The sub-pixel shift dual-energy

CBCT data acquisition scheme is proposed for DL-FPD for the ﬁrst time. This method

perfectly ﬁts with the stacked structure of the DL-FPD to increase the spatial sampling rate

while shortening the signal readout time by introducing additional “geometric mismatch”

between the two detector layers. (2) A high performance one-step material decomposition

algorithm is developed for the proposed novel DL-FPD based sub-pixel shift dual-energy

CBCT imaging. The transmitted X-ray intensity of the sub-ray in each sub-pixel together

with the CBCT imaging geometry, the data noise statistics and the beam spectra are utilized

to build a more accurate forward imaging model, and the penalized likelihood function is

optimized to generate the basis images with high spatial resolution.

The rest of this paper is organized below: Section II introduces some related works. Sec-

tion III presents the mathematical model of the sub-pixel shift data acquisition method for

DL-FPD, the penalized likelihood material decomposition method, the experimental setup,

the implementation details and evaluation metrics. Section IV presents the experimental

results of diﬀerent objects. Section Vprovides the discussions and a brief conclusion.

II. RELATED WORK

A. Super resolution CT imaging method

There are two main strategies to realize super spatial resolution imaging in medical CBCT

applications. First, the ﬂying focal spot (FFS) technique10,11 used by some advanced X-ray

tube doubles the spatial sampling density in the horizontal and vertical directions, and

thus can signiﬁcantly reduce the in-plane and the out-of-plane aliasing artifacts. Since the

detector matrix size is ﬁxed, the FFS technique usually augments the total data size and

adds burdens to the detector read-out speed. Second, the sub-pixel shifting technique in

the detector end, which has been proposed for optical imaging for a long time12, and can

also be utilized to greatly improve the spatial resolution of CT images. For example, Yan

et al. proposed to shift the single-layer detector array and make multiple scans to jointly

reconstruct the high spatial resolution CT image13. Li et al. suggested to move the rotation

stage in a ﬁxed trajectory during the data acquisition to obtain multiple projection images

with sub-pixel displacements14. Szczykutowicz et al. developed a spectral-spatial encoding

method that acquires multiple measurements at diﬀerent beam energies and diﬀerent spa-

tial positions to decompose the line integrals into two unique bases15. However, one major

limitation of these studies is the prolonged data acquisition period due to the repeated mea-

surements. In this work, the proposed suRi method can eﬃciently overcome such limitation

by incorporating the sub-pixel shifts into the DL-FPD.

文档加载中……请稍候！
如果长时间未打开，您也可以点击刷新试试。

下载文档到电脑，查找使用更方便

10 玖币 0人已下载

立即下载

摘要：

Superresolutiondual-energycone-beamCTimagingwithdual-layerat-paneldetectorTingSu,1JiongtaoZhu,2XinZhang,1DongZeng,3YuhangTan,1HanCui,1HairongZheng,4JianhuaMa,3DongLiang,1,4andYongshuaiGe1,4,a)1)ResearchCenterforMedicalArticialIntelligence,ShenzhenInstitutesofAdvancedTechnology,ChineseAcademyofScie...

展开>> 收起<<

Super resolution dual-energy cone-beam CT imaging with dual-layer ﬂat-panel detector Ting Su1Jiongtao Zhu2Xin Zhang1Dong Zeng3Yuhang Tan1Han Cui1Hairong.pdf

共30页,预览5页

还剩页未读，继续阅读

声明：本站为文档C2C交易模式，即用户上传的文档直接被用户下载，本站只是中间服务平台，本站所有文档下载所得的收益归上传人(含作者)所有。玖贝云文库仅提供信息存储空间，仅对用户上传内容的表现方式做保护处理，对上载内容本身不做任何修改或编辑。若文档所含内容侵犯了您的版权或隐私，请立即通知玖贝云文库，我们立即给予删除！

Super resolution dual-energy cone-beam CT imaging with dual-layer ﬂat-panel detector Ting Su1Jiongtao Zhu2Xin Zhang1Dong Zeng3Yuhang Tan1Han Cui1Hairong

相关推荐

开通VIP享超值会员特权

作者详情

相关内容

热门标签

举报选择: