Nasopharyngeal-carcinoma (NPC) lesion automatic-segmentation method and nasopharyngeal-carcinoma lesion automatic-segmentation systems based on deep learning

Abstract

The invention discloses a nasopharyngeal-carcinoma (NPC) lesion automatic-segmentation method and nasopharyngeal-carcinoma lesion automatic-segmentation systems based on deep learning. The method comprises: carrying out registration on a PET (Positron Emission Tomography) image and a CT (Computed Tomography) image of nasopharyngeal carcinoma to obtain a PET image and a CT image after registration;and inputting the PET image and the CT image after registration into a convolutional neural network to carry out feature representation and scores map reconstruction to obtain a nasopharyngeal-carcinoma lesion segmentation result graph. The method carries out registration on the PET image and the CT image of the nasopharyngeal carcinoma, obtains a nasopharyngeal-carcinoma lesion by automatic segmentation through the convolutional neural network, and is more objective and accurate as compared with manual segmentation manners of doctors; and the convolutional neural network in deep learning isadopted, consistency is better, feature learning ability is higher, the problems of dimension disasters, easy falling into a local optimum and the like are solved, lesion segmentation can be carried out on multi-modal images of the PET-CT images, and an application range is wider. The method can be widely applied to the field of medical image processing.

Description

Technical field [0001] The invention relates to the field of medical image processing, in particular to a method and system for automatic segmentation of nasopharyngeal carcinoma lesions based on deep learning. Background technique [0002] The imaging principle of positron emission tomography (Positron Emission Tomography, PET) is to label compounds that can participate in human metabolism with radionuclides. The synthesized substances are called imaging agents or tracers. Considering that large radiation doses are harmful to human health, etc. Factors, generally use short half-life radionuclides, such as: 18F labeling glucose, 11C labeling choline, 13N labeling amino acids, etc. By injecting these tracers into the subject, they can participate in the subject's metabolic process. When radionuclides are involved in metabolism, they decay at the same time. Protons release positrons and neutrinos to decay into neutrons. After moving about 1-3mm in the human body, the positrons com...

AI Technical Summary

Problems solved by technology

However, the disadvantage of PET images is that they cannot provide clear human anatomy and cannot make accurate diagnosis

However, limited by traditional machine learning methods, these methods have shortcomings such as insufficient feature learning ability, dimensionality disaster, and easy to fall into local optimum, and they are only suitable for lesion segmentation of single-modal images such as PET images or CT images, not suitable for The lesion segmentation of PET-CT images, a multimodal image, needs to be further improved

Images