A medical image classifier is proposed in this paper.Gradient Vector Flow (GVF) is used to segment tumor area
and the texture and shape features are the inputs to classifier.A novel SVM algorithm is presented in this paper
which solves the problem that SVM has a poor performance when the two-class problem samples are very unbalanced.The experimental results indicate that the modified algorithm has a strong capability of classification for the unbalanced samples
and the efficiency and accuracy in the medical images are greatly improved.