Small intrapulmonary nodule progress evaluation system and method based on deep learning

Abstract

The invention provides a small intrapulmonary nodule progress evaluation system and method based on deep learning, and relates to the field of medical image processing technology. The method comprisesthe steps of acquiring all lung tomography images in a follow-up process of a small intrapulmonary nodule patient; performing image preprocessing on all the lung tomography images for obtaining preprocessed images; for each small intrapulmonary nodule patient, marking the small intrapulmonary nodules at the same position in each preprocessed image and establishing a nodule growth database; establishing an initial evaluation model according to a preset hyperparameter; training the initial evaluation model accordingi to the nodule growth database for obtaining a progress evaluation model; and inputting the lung tomography images of a to-be-evaluated small intrapulmonary nodule patient into the progress evaluation model at different follow-up time for obtaining growth progress data of the small intrapulmonary nodule of the small intrapulmonary nodule patient, thereby supplying to a doctor for diagnosis reference. The method of the invention can objectively and accurately evaluate changesof the small intrapulmonary nodule and effectively improves evaluation result accuracy.

Description

technical field [0001] The present invention relates to the technical field of medical image processing, in particular to a system and method for assessing the progress of small pulmonary nodules based on deep learning. Background technique [0002] With the popularization of cancer "early diagnosis and early treatment" awareness and the improvement of tomographic imaging technology, especially the application and popularization of thin-section high-resolution tomography, more and more pulmonary nodules have begun to be discovered, among which small pulmonary nodules are the most common. How to evaluate the progress of small pulmonary nodules found by thin-section tomography is still a clinical problem. It is often difficult to characterize small pulmonary nodules after they are discovered, and a period of follow-up and detection is required to help formulate a follow-up treatment plan based on the growth characteristics and progress of the small nodules. Clinically, for p...

AI Technical Summary

Problems solved by technology

In the tomographic image analysis of pulmonary lesions, the existing technology still focuses on the detection and quality determination of pulmonary nodules, while ignoring the application in the dynamic monitoring of pulmonary nodules

Images