CTSim: a numerical simulator of computed tomography for high-quality radiological education

Int J Comput Assist Radiol Surg. 2022 Jul;17(7):1257-1269. doi: 10.1007/s11548-022-02656-6. Epub 2022 May 27.

Abstract

Purpose: Computer-based simulation offers radiological students the possibility to experiment with computed tomography in a way not possible in class or in clinical practice. The aim of this study was to design a computed tomography numerical simulator (CTSim) for high-quality radiological education.

Methods: In this study, a CTSim is designed based on the mathematical and physical principles of CT imaging. The proposed CTSim includes pen-beam module, fan-beam module, and clinical CT module. The core design of the software includes four parts: the construction of sample models, construction of imaging parameters and artifact parameters, design of data acquisition models under different scanning modes, and design of image reconstruction algorithm. After the design of the CTSim, the proposed CTSim was tested in every step of CT imaging.

Results: Systematic verification demonstrated that the proposed CTSim can not only perform raw CT data acquisition, image reconstruction, basic image processing, and image quality analysis like a real CT scanner, but can also simulate the formation of artifacts. The CTSim can completely get rid of the hardware and achieve the same experimental results as the hardware instrument.

Conclusion: The proposed CTSim software shows several advantages such as low cost, less room accommodation, and no ionizing radiation damage and can be used as a virtual experimental training platform to enhance teaching and learning for general X-ray CT courses or for self-study of CT practitioners.

Keywords: Artifact analysis; CT imaging principle simulation; Dynamic data acquisition; Image reconstruction; Numerical computation simulation.

MeSH terms

  • Algorithms
  • Artifacts
  • Humans
  • Image Processing, Computer-Assisted* / methods
  • Phantoms, Imaging
  • Tomography, X-Ray Computed* / methods