Abstract: Medical image fusion can integrate different facilities image diagnoses formation,takes full advantage of various facilities merit.It not only enhance disease accuracy but also improves focus precise location,and has a very important value of application for diagnosis.Wavelet transform is a effective method of image fusion.This paper utilizes biorthogonal wavelet transform merge CT and PET image for avoiding phase distortion by conventional orthogonal wavelet transform,and introduces a new fusion method.Experimental result shows this method can effectively integrate anatomic and functional information,preserves the details of each input image successfully.

摘要： 医学图像融合能将不同设备图像的诊断信息综合起来,充分利用各成像设备的优点。不仅可以提高疾病的确诊率,而且可以提高病灶的定位精度,在诊断上具有重要的应用价值。小波变换是图像融合中一种有效的方法。为了避免正交小波产生的相位失真,采用双正交小波对CT-PET图像进行多分辨率分解与重构,并提出一种新的融合算法。实验表明,该方法能有效地将解剖信息和功能信息集成在一起,完好地保留了各图像的信息。