Method for constructing three-dimensional medical image

A method for constructing three-dimensional medical image comprising steps of:

• inputting multiple two-dimensional images and a known three-dimensional image into a processing module, wherein:
• the two-dimensional images comprise multiple characteristic parameters corresponding to a specific body part and the known three-dimensional image comprises an image of the specific body part, wherein an X-ray module generates the two-dimensional images;
• and
• the processing module generates a neural network through a deep learning algorithm with reference to the known three-dimensional image and unfolding the two-dimensional images to produce a three-dimensional reconstruction;
• and
• inputting a new two-dimensional image into the processing module to obtain a reconstructed three-dimensional image, wherein the processing module utilizes the neural network to build the reconstructed three-dimensional image by unfolding the new two-dimensional image to produce the three-dimensional reconstruction.