Development of a prototype chest digital tomosynthesis (CDT) R/F system with fast image reconstruction using graphics processing unit (GPU) programming

https://doi.org/10.1016/j.nima.2016.12.027Get rights and content

Abstract

Digital tomosynthesis offers the advantage of low radiation doses compared to conventional computed tomography (CT) by utilizing small numbers of projections (~80) acquired over a limited angular range. It produces 3D volumetric data, although there are artifacts due to incomplete sampling. Based upon these characteristics, we developed a prototype digital tomosynthesis R/F system for applications in chest imaging. Our prototype chest digital tomosynthesis (CDT) R/F system contains an X-ray tube with high power R/F pulse generator, flat-panel detector, R/F table, electromechanical radiographic subsystems including a precise motor controller, and a reconstruction server. For image reconstruction, users select between analytic and iterative reconstruction methods. Our reconstructed images of Catphan700 and LUNGMAN phantoms clearly and rapidly described the internal structures of phantoms using graphics processing unit (GPU) programming. Contrast-to-noise ratio (CNR) values of the CTP682 module of Catphan700 were higher in images using a simultaneous algebraic reconstruction technique (SART) than in those using filtered back-projection (FBP) for all materials by factors of 2.60, 3.78, 5.50, 2.30, 3.70, and 2.52 for air, lung foam, low density polyethylene (LDPE), Delrin® (acetal homopolymer resin), bone 50% (hydroxyapatite), and Teflon, respectively. Total elapsed times for producing 3D volume were 2.92 s and 86.29 s on average for FBP and SART (20 iterations), respectively. The times required for reconstruction were clinically feasible. Moreover, the total radiation dose from our system (5.68 mGy) was lower than that of conventional chest CT scan. Consequently, our prototype tomosynthesis R/F system represents an important advance in digital tomosynthesis applications.

Introduction

According to the American Cancer Society, lung cancer was the second most common cancer for both men and women in the United States in 2015, followed by prostate cancer and breast cancer, respectively [1]. Accurate diagnosis is required for early detection and optimal treatment of lung cancer. Chest radiography (CXR) has been considered the most basic type of study for pulmonary disease because of its cost effectiveness and accessibility. However, solitary lung nodule detection using CXR is often challenging because 2D chest projections may partially or completely obscure pulmonary nodules due to superposition of internal structures [2], [3].

Digital tomosynthesis has been considered an attractive modality because it can address the problem of poor depth resolution in conventional digital radiography [4]. It is also cost effective because it requires only simple hardware, an X-ray tube, and a detector, without a robotic gantry. Several prototype and commercial tomosynthesis systems have been developed for clinical applications in small lung nodules to deliver much lower radiation doses than computed tomography (CT) [5], [6], [7], [8], [9], [10].

Based upon these advantages, we have developed a prototype digital tomosynthesis R/F system that is applicable for chest imaging. Although there has been high interest in developing tomosynthesis, most systems are distributed for breast tomosynthesis. Only two commercially released chest tomosynthesis products have recently been approved by the Food and Drug Administration (FDA) [7], [11]. Chest digital tomosynthesis (CDT) often demands fast scan time of less than 10–15 s to produce 70–80 X-ray images. Therefore, the development of CDT systems requires precise geometric motion control of the X-ray tube and detector. Shan et al. described mechanical challenges secondary to the requirement for fast scanning times, which may lead to X-ray focal spot blur during image acquisition [5]. Another major difficulty of tomosynthesis is that it takes considerable time (~5 mins) to reconstruct 3D volumes, although the exact time depends on the amount of data to be processed.

In this study, the development and operation of a prototype CDT R/F system was described. The purpose of this study is to investigate the feasibility of our newly developed CDT system to contribute to tomosynthesis technology. The prototype system was operated by carefully controlling its electromechanical subsystems through a synchronized interface, including a tube rotator with a 6-axis motor board and a high-frequency generator (kVp, mA, exposure time with R/F pulse). To address the problem of extended reconstruction time, graphics processing unit (GPU) programming was used to accelerate the reconstruction algorithm for both analytic and iterative reconstruction methods. A phantom study was conducted to obtain and analyze tomosynthesis images acquired with the system and to validate its utility, with favorable results.

Section snippets

Prototype CDT system

A prototype CDT system was developed and 41 projection images were acquired over a sweep angle range of ±20° with 1° angle steps. A schematic block diagram of the prototype system is shown in Fig. 1. This system consists of an X-ray tube (E7869X, Toshiba), a CsI scintillator flat panel detector (FPD) (Pixium RF 4343, Thales), a R/F table, a main controller (remote user console, main control board, and command processor), and a reconstruction server. The main controller controls a high power

Results

We developed a prototype CDT R/F system and operated it for clinical applications. We designed a control board using FPGA to synchronize the interface signal from the touch console, X-ray tube and detector, and reconstruction server. The results indicate that our system was successfully operated with recorded geometric positions from the encoding driver software. Moreover, the high powered X-ray R/F pulse showed rapid rise time and fall time during intervals from charge state to discharge state

Discussion and conclusions

We successfully developed and operated an efficient and cost effective prototype CDT R/F system. We utilized distance-driven forward/backward projection as a reconstruction strategy for accurate modeling in our CDT system. Distance-driven schemes usually require long computation times compared to conventional pixel- or ray-driven methods [14]. We assigned a thread to each pixel (forward projection) or each voxel (back-projection) to calculate weighting coefficients. The advantage of fast

Acknowledgments

This research was financially supported by the Ministry of Trade, Industry & Energy (MOTIE), The Korea Institute for Advancement of Technology (KIAT), and the Gangwon Institute for Regional Program Evaluation (GWIRPE) through the Encouragement Program for The Industries of Economic Cooperation Region (No. R0002898).

References (24)

  • B.G. Brogdon et al.

    Radiol. Clin. N. Am.

    (1983)
  • J.T. Dobbins et al.

    Eur. J. Radiol.

    (2009)
  • A. Andersen et al.

    Ultrason. Imaging

    (1984)
  • D. Van de Sompel et al.

    Med. Image Anal.

    (2011)
  • R.L. Siegel, K.D. Miller, A. Jemal, CA: A Cancer Journal For Clinicians, vol 65, 1, 2015, pp....
  • P.K. Shah et al.

    Radiology

    (2003)
  • J.T. Dobbins et al.

    Phys. Med. Biol.

    (2003)
  • J. Shan et al.

    Phys. Med. Biol.

    (2015)
  • J. Vikgren et al.

    Radiology

    (2008)
  • J.T. Dobbins et al.

    Med. Phys.

    (2008)
  • Y. Yamada et al.

    Invest. Radiol.

    (2011)
  • Y. Yamada et al.

    Eur. Radiol.

    (2013)
  • Cited by (9)

    • Comparison study of various beam modulation schemes on image quality in chest digital tomosynthesis (CDT) system

      2018, Radiation Physics and Chemistry
      Citation Excerpt :

      However, the images acquired from 2D digital radiography do not present depth information, and this technique is limited in regard to accurate diagnosis when lung nodule overlap with the internal structure. Recently, chest digital tomosynthesis (CDT) systems have been actively developed and studied to overcome these limitations using a smaller dose than that of computed tomography (CT) systems (Choi, 2017). However, the CDT system requires a higher exposure dose than 2D digital radiography systems, and the results of many studies suggest that radiation dose reduction methods for the CDT systems are required (Svalkvist and Bath, 2010; Johnsson, 2014).

    • Interactive Garden Landscape Digital Reconstruction System Based on Particle Swarm Algorithm

      2022, Lecture Notes on Data Engineering and Communications Technologies
    • Application of nonlinear voxel distribution grid for computational speed-up for linear tomosynthesis reconstruction

      2018, Proceedings of SPIE - The International Society for Optical Engineering
    View all citing articles on Scopus
    View full text