The invention relates to the technical field of medical CT image processing, and discloses a CT image segmentation method based on an artificial neural network, and the method comprises the following steps: carrying out the preprocessing of a CT image; segmenting the skeleton part, obtaining the outer contour of the abdominal cavity, and determining the number of faults; carrying out internal organ feature recognition in the outline of the skeleton, and extracting the outline of a single internal organ; processing, segmenting and storing the outline of the viscera in a classified mode; judging whether all viscera organs in the fault are segmented or not, if not, returning to reprocess, and otherwise, performing the next step; extracting a plurality of pieces of fault data of a certain viscera organ in a centralized manner according to requirements; According to the method, the skeleton part is firstly segmented, the area where the internal organs are located is rapidly determined, data processing of follow-up internal organ recognition and segmentation is reduced, invalid recognition is reduced, and the segmentation speed is increased; by setting the standard library, the segmented image data is detected, and the integrity of the segmented image is ensured.