Automatic extraction based on heart bottom ct nonlinear gradation image Model

- bottomed cardiac CT image based on the gradation nonlinear model automatic extraction method, comprising the steps of: acquiring a sequence of cardiac CT; respectively calculating a first characteristic value of each layer in the sequence of cardiac CT lung regions; a first characteristic value using the heart Construction top and bottom constraints of the heart, solving the constraint condition into the sequence of cardiac CT sequence top and bottom sequence; sequence, respectively, of the top and the bottom layer extracted cardiac apex disappears valid area information; using the extracted linear gradient model of the heart cardiac bottom structure effective region information, the edge profile to fit the heart of the base sequence of the layers, thereby obtaining an accurate bottom heart. The method of the present invention can accurately extracted at the bottom of the heart and chest diaphragm adhesions. And robust, high operating efficiency, low hardware configuration requirements for network applications.