eduzhai > Applied Sciences > Engineering >

A novel deep learning-based method for monochromatic image synthesis from spectral CT using photon-counting detectors

  • king
  • (0) Download
  • 20210506
  • Save

... pages left unread,continue reading

Document pages: 9 pages

Abstract: With the growing technology of photon-counting detectors (PCD), spectral CTis a widely concerned topic which has the potential of materialdifferentiation. However, due to some non-ideal factors such as cross talk andpulse pile-up of the detectors, direct reconstruction from detected spectrumwithout any corrections will get a wrong result. Conventional methods try tomodel these factors using calibration and make corrections accordingly, butdepend on the preciseness of the model. To solve this problem, in this paper,we proposed a novel deep learning-based monochromatic image synthesis methodworking in sinogram domain. Different from previous deep learning-based methodsaimed at this problem, we designed a novel network architecture according tothe physical model of cross talk, and it can solve this problem better in aningenious way. Our method was tested on a cone-beam CT (CBCT) system equippedwith a PCD. After using FDK algorithm on the corrected projection, we got quitemore accurate results with less noise, which showed the feasibility ofmonochromatic image synthesis by our method.

Please select stars to rate!

         

0 comments Sign in to leave a comment.

    Data loading, please wait...
×