Oral semantic model training method and oral cone beam CT image optimization method and device

• Main Authors: WANG LEI, ZHOU SIYI, CHEN LEI, LIU ZHENG, CHEN YIPING, JIA KAIYUE, LI MIN, HUANG WEI, CHEN RAN
• Format: Patent
• Language: Chinese; English
• Published: 08.09.2023
• Subjects: CALCULATING; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS; COMPUTING; COUNTING; PHYSICS

The invention discloses an oral semantic model training method and an oral cone beam CT image optimization method and device, and belongs to the technical field of medical imaging, and the method comprises the steps: obtaining a plurality of continuous first CBCT images and a plurality of continuous second CBCT images from a preset two-dimensional CBCT image set; obtaining a corresponding value segmentation image according to the second CBCT image; obtaining an invariant extraction feature in the first CBCT image so as to reduce a marking range of an oral semantic model on a value segmentation map corresponding to the second CBCT image; and inputting the value segmentation image corresponding to the second CBCT image into the oral semantic model until the output model loss value is no longer reduced. According to the method and the device, the plurality of continuous first CBCT images and the plurality of continuous second CBCT images are acquired, and the corresponding value segmentation map is acquired acco