3D　shape　reconstruction　of　the　liver　from　its　2D　cross-sections　improves　the　surgeon＇s　knowledge　of　liver　anatomy　and　makes　even　more　complicated　liver　surgery　safe,　which　can　be　employed　to　aid　clinical　practice　as　an　alternative　tool.　In　this　paper,　a　new　method　for　3D　reconstruction　of　liver　is　proposed.　It　mainly　consists　of　three　steps:　liver　auto-segmentation,　surface　reconstruction　and　surface　rendering.　Firstly,　an　effective　automatic　segmentation　method　based　on　graph-theory　is　proposed　for　extracting　liver.　Then,　NMC　algorithm　based　on　a　combination　of　MC　and　Cuberille　algorithms　is　applied　to　accomplish　surface　reconstruction　of　liver.　Finally,　surface　rendering　of　liver　is　implemented　by　3D　graphics　library　OpenGL.　The　method　mentioned　above　is　being　tested　on　numerous　experiments　of　3D　reconstruction　of　liver　and　results　are　promising.