Speaker
Description
X-ray computed tomography (CT) based on GOS scintillator coupled with photodiode is widely used in medical imaging. However, high image quality in conventional CT requires a high radiation dose, which leads to an inability to acquire energy information of X-rays. Therefore, as the next generation CT, the photon-counting CT (PCCT) has been proposed, which provides multi-energy and low-radiation-dose imaging. In PCCT, CT images in different energy bands can be subtracted to extract specific materials, a process known as “K-edge-imaging”. However, whether the obtained PCCT images corresponds to individual energy bands, as expected, has not been clearly investigated. Some previous studies have suggested that the energy information in PCCT, specifically low-energy information, may be substantially inaccurate due to the contamination of escape or Compton scattered events. In this study, we investigated the accuracy of energy information in obtained PCCT images. The PCCT detector developed in our system consisted of a multi-pixel photon counter coupled with a fast ceramic YGAG:Ce scintillator, thus can be easily constructed, has a lower total cost and has high compatibility with conventional CT systems. By imaging various phantoms, we confirmed that the experimental CT values in the low-energy bands is actually deviated from the theoretical values. Based on these results, we developed a novel correction method to reduce the effects of energy-miscounted X-ray photons by using the Geant4 simulation. As a result, the contrast of the CT images and CT values in the low-energy bands improved by approximately 35% after applying the correction. This correction method is independent of the phantom properties, which means that it can be applied in various imaging cases. We conclude that the proposed method can be effective at improving the accuracy of low-energy information in PCCT, which expands the application of PCCT energy information for clinical uses.
| Do you need a VISA letter for traveling to Canada ? | No |
|---|
