Objective: This article aimed to achieved a surgery planning that used digital medical technology to repair a transmedian mandibular defect, and designed the digital model of titanium plate for mandibular reconstruction.Also it showed the related technical process, hoping to give some idea for clinical workers.Methods: The patient is a 45-year-old female diagnosed with mandibular ameloblastoma.The CT examination data of mandible with the slice thickness of lmm was taken.The CT data was imported into the software of Mimics 10.01 (Co.Materialise, Belgium) for image segmentation and 3-d reconstruction.They used the function of simulation to cut off the tumor and caused an 83 mm long defect area.They used the function of mirror image to repair the defect area.After that, a symmetrical defect area of 13 mm nears the median still remains (see figure 1).They import the data into the software of Matlab 2010b (Co.MathWorks, American).They used the algorithm of support vector machine (SVM), one of the machine learning (ML) technologies which can calculate any defect area data according to the remaining data of mandible, to repair the defect area(see figure 2).