A postreconstruction method for correcting the beam-hardening artifacts in computed tomography (CT) images is proposed. This method does not require x-ray spectrum measurement. The authors assumed that a pixel in a CT image can be decomposed into equivalent tissue percentages, depending on its CT number. A scout view of the step wedges made of these equivalent tissues was performed to obtain a beam-hardening correction curve for each tissue. Projecting through the CT image from various angles generated simulated projection data and the total thickness of each tissue along the ray. The correction term was estimated using the tissue thickness traveled by the ray, and this term was then added to its corresponding projection data. A second reconstruction using the corrected projection data yielded a beam-hardening corrected image. The preliminary results show that this method reduces beam hardening artifacts by 14% for aluminum and increased the object contrast by 18% near the aluminum-water boundary. The variation in CT numbers at different locations were reduced, and the aluminum CT number also was restored.
Purpose:
In positron emission tomography (PET), the single scatter simulation (SSS) algorithm is widely used for scatter estimation in clinical scans. However, bias usually occurs at the essential steps of scaling the computed SSS distribution to real scatter amounts by employing the scatter‐only projection tail. The bias can be amplified when the scatter‐only projection tail is too small, resulting in incorrect scatter correction. To this end, we propose a novel scatter calibration technique to accurately estimate the amount of scatter using pre‐determined scatter fraction (SF) function instead of the employment of scatter‐only tail information.
Methods:
As the SF depends on the radioactivity distribution and the attenuating material of the patient, an accurate theoretical relation cannot be devised. Instead, we constructed an empirical transformation function between SFs and average attenuation coefficients based on a serious of phantom studies with different sizes and materials. From the average attenuation coefficient, the predicted SFs were calculated using empirical transformation function. Hence, real scatter amount can be obtained by scaling the SSS distribution with the predicted SFs. The simulation was conducted using the SimSET. The Siemens Biograph™ 6 PET scanner was modeled in this study. The Software for Tomographic Image Reconstruction (STIR) was employed to estimate the scatter and reconstruct images. The EEC phantom was adopted to evaluate the performance of our proposed technique.
Results:
The scatter‐corrected image of our method demonstrated improved image contrast over that of SSS. For our technique and SSS of the reconstructed images, the normalized standard deviation were 0.053 and 0.182, respectively; the root mean squared errors were 11.852 and 13.767, respectively.
Conclusion:
We have proposed an alternative method to calibrate SSS (C‐SSS) to the absolute scatter amounts using SF. This method can avoid the bias caused by the insufficient tail information and therefore improve the accuracy of scatter estimation.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.