Jirasek A, Matthews Q, Hilts M, Schulze G, Blades M W, Turner R F B
Department of Physics and Astronomy, The University of Victoria, Victoria BC V8W 3P6, Canada.
Phys Med Biol. 2006 May 21;51(10):2599-617. doi: 10.1088/0031-9155/51/10/016. Epub 2006 May 4.
This study presents a new method of image signal-to-noise ratio (SNR) enhancement by utilizing a newly developed 2D two-point maximum entropy regularization method (TPMEM). When utilized as an image filter, it is shown that 2D TPMEM offers unsurpassed flexibility in its ability to balance the complementary requirements of image smoothness and fidelity. The technique is evaluated for use in the enhancement of x-ray computed tomography (CT) images of irradiated polymer gels used in radiation dosimetry. We utilize a range of statistical parameters (e.g. root-mean square error, correlation coefficient, error histograms, Fourier data) to characterize the performance of TPMEM applied to a series of synthetic images of varying initial SNR. These images are designed to mimic a range of dose intensity patterns that would occur in x-ray CT polymer gel radiation dosimetry. Analysis is extended to a CT image of a polymer gel dosimeter irradiated with a stereotactic radiation therapy dose distribution. Results indicate that TPMEM performs strikingly well on radiation dosimetry data, significantly enhancing the SNR of noise-corrupted images (SNR enhancement factors >15 are possible) while minimally distorting the original image detail (as shown by the error histograms and Fourier data). It is also noted that application of this new TPMEM filter is not restricted exclusively to x-ray CT polymer gel dosimetry image data but can in future be extended to a wide range of radiation dosimetry data.
本研究提出了一种利用新开发的二维两点最大熵正则化方法（TPMEM）增强图像信噪比（SNR）的新方法。当用作图像滤波器时，二维TPMEM在平衡图像平滑度和保真度的互补要求方面展现出无与伦比的灵活性。该技术被评估用于增强辐射剂量测定中使用的辐照聚合物凝胶的X射线计算机断层扫描（CT）图像。我们利用一系列统计参数（例如均方根误差、相关系数、误差直方图、傅里叶数据）来表征应用于一系列具有不同初始SNR的合成图像的TPMEM的性能。这些图像旨在模拟X射线CT聚合物凝胶辐射剂量测定中会出现的一系列剂量强度模式。分析扩展到用立体定向放射治疗剂量分布辐照的聚合物凝胶剂量计的CT图像。结果表明，TPMEM在辐射剂量测定数据上表现出色，能显著提高噪声干扰图像的SNR（SNR增强因子>15是可能的），同时对原始图像细节的扭曲最小（如误差直方图和傅里叶数据所示）。还指出，这种新的TPMEM滤波器的应用不仅限于X射线CT聚合物凝胶剂量测定图像数据，未来还可扩展到广泛的辐射剂量测定数据。
