Whole heart ct segmentation method and device based on multiple labels and multiple decoders

The invention provides a whole heart ct segmentation method and device based on multiple labels and multiple decoders, and the method comprises the steps: obtaining a training CT image which comprises multiple to-be-segmented target objects and first label information corresponding to the multiple to-be-segmented target objects; the first label information of the multiple to-be-segmented target objects is fused according to a preset rule, multiple pieces of second label information are obtained, and the granularity of the first label information is smaller than that of the second label information; preprocessing the CT image, and cutting the preprocessed CT image into a plurality of image blocks with preset sizes; a U-net network is constructed, the U-net network comprises a first decoder and a second decoder, the first decoder is used for segmenting fine-grained labels, and the second decoder is used for segmenting coarse-grained labels; according to the image blocks, the first label information and the seco