基于自适应非线性拉伸的DR胸片图像处理

目的本文首先对目前的DR图像处理技术进行总结,并分析胸片图像的特殊性,针对目前图像处理算法对胸片图像处理的不足,提出了一种自适应非线性拉伸的DR胸片图像处理算法。方法该算法在对图像进行增强、均衡、去噪之后,根据图像的统计信息确定感兴趣区,然后利用非线性变换LUT表对图像进行变换。结果在实验部分对Hologic DR探测器产生的图像进行处理,胸片图像细节突出。结论结果表明应用本文算法进行处理后,胸片的感兴趣区对比度增加,细节信息更丰富,更有利于医生进行诊断。

Processing of Chest X-ray Digital Radiography Images Based on Auto Adaptive Nonlinear Stretching

Objective To summarize the current processing technologies of digital radiography（DR） images, analyze the specificity of chest X-ray images,improve the drawbacks of current image processing algorithms,and introduce a novel image processing algorithm based on auto adaptive nonlinear stretching method. Methods After enhancing,balancing and denoising,the images were transformed by a LUT table of nonlinear transform, according to the statistics information of region of interesting（ROI）. Results In the experiment,chest X-ray images gained from the Hologic DR detector were processed and details in the images were strengthened. Conclusion The contrast of ROI of chest X-ray images enhanced and more detail information appeared after procession of algorithm. The processed images was more convenient for doctors to make diagnosis on those processed images.