Coronary angiography three-dimensional reconstruction method based on neural network radiation field

• Main Authors: WANG FEIFAN, MA JUN, LAN HONGZHI, ZHENG LINGXIAO
• Format: Patent
• Language: Chinese; English
• Published: 03.05.2024
• Subjects: CALCULATING; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS; COMPUTING; COUNTING; IMAGE DATA PROCESSING OR GENERATION, IN GENERAL; PHYSICS

The invention provides a coronary angiography three-dimensional reconstruction method based on a neural network radiation field. The coronary angiography three-dimensional reconstruction method comprises the steps that a single coronary angiography image, collected by an angiography camera, of a target coronary artery is acquired; segmenting the single coronary angiography image to obtain a coronary segmented image; inputting the coronary artery segmentation image into a pre-trained feature extraction network for feature extraction to obtain image features of the coronary artery segmentation image; inputting the image features and visual angle information of the radiography camera when the radiography camera collects a single coronary angiography image into a pre-trained neural network radiation field to obtain transparency data; and reconstructing a three-dimensional model of the target coronary artery according to the transparency data. Therefore, three-dimensional model reconstruction of the coronary arter