【摘要】 为乳腺癌早期诊断和乳腺X线影像微钙化点计算机辅助检测作前期预处理,提出了一种基于小波变换的微钙化点感兴趣区提取新技术。其具体思路是:(1)将乳腺区域图像提取成等大的子图像;(2 )对每一幅子图像进行小波变换,根据特征参数ρ讨论最优小波变化参数和阈值T;(3)根据阈值T判别子图像是否属于感兴趣区。对临床实际病例(2 0幅乳腺X线影像)的试验结果表明,该方法具有较高的检出率(89.7% ) ,和较为满意的假阳性率(2 .1% )。更多还原

【Abstract】 In order to preprocess mammograms for diagnosing the early cases of breast cancer and improving the computational efficiency in the comput er-aided detection of micro-ca l cifications in mammograms,we have advanced a no vel processing technique for the extraction of micro-calcification region of interest (MROI). The proposed method is based on a three-step procedu r e: (1) the mammogram is divided into sub-images of the same size; (2) the wavele t multi-resolution method is conducted on the sub-images, and the parameters r elated to wavelet transform and threshold T are discussed according to ρ; (3) the classification of sub-images is determined by T. It is tested w it h 20 mammograms and the results show that the method can achieve a true positiv e rate as high as 89.7% with a false positive rate as low as 2.1%.更多还原