基于阈值的平板探测器增益校正方法

A Gain Correction Method Based on Threshold for Flat-Panel Detector

为了更好地提高平板探测器在齿科CT中采集图像的质量,去除由于探测器制造缺陷等因素而导致的CT重建切片中的条状和环状伪影,提出了一种基于阈值的平板探测器增益校正方法.利用采集的相同放射条件下的空气数据,通过归一化方法计算得到增益系数矩阵.考虑到物体对射线的衰减性,先通过阈值区分出图像的不同部分,然后根据增益系数矩阵和阈值判断结果,对图像的每个像素逐一进行校正.实验结果表明:在校正后的图像中,由于探测器制造缺陷等因素所引起的重建切片中的大量条纹以及环状伪影被有效地清除,且未出现新引入的条纹,重建切片的图像质量得到明显改善.

In order to remove the stripes and ring artifacts caused by the defects of the detector and to improve the reconstructed image quality from flat panel detector in dental CT, a new gain correction method based on threshold was proposed. The air images were collected under the same radiation conditions as the normal scan to calculate the gain coefficient matrix by the normalization method. What＇ s more, the attenuation of the object was taken into account. With the matrix of the gain coefficient and the result of the threshold judgment, the gain correction was performed on each pixel of the image one by one according to the calculated gain coefficient matrix. Experimental result showed that a large number of stripes and ring artifacts in the image were effectively removed after the correction, with no stripes introduced again. The quality of the reconstructed image was improved significantly.