Feature extraction is an important part in Content-based image retrieval (CBIR).It is an active research area over the past few decades. In this paper texture feature extraction of mammogram images are done. Biorthogonal wavelet filter via lifting scheme is used for the extraction of texture features. Maximum likelihood estimator (MLE) is used for texture feature estimation. Here Digital Database for Screening Mammography (DDSM) is used as the database. Here biorthogonal wavelets are used in the lifting scheme to get texture feature vectors of mammogram images. By using lifting scheme in all biorthogonal wavelets, predict and update filter coefficients are also got. These coefficients will be adapted later and thus we can find the optimal wavelet filter bank for increasing the retrieval performance of the retrieval system. By using lifting scheme methodology decomposition of images are done and thus got approximation and detail coefficients of image.