Nasopharyngeal carcinoma positioning segmentation method and system based on image segmentation convolutional neural network

Abstract

The invention discloses a nasopharyngeal carcinoma positioning segmentation method and system based on an image segmentation convolutional neural network, and the method comprises the steps: obtaining an electronic nose endoscopic image through employing a WLI mode and an NBI mode, inputting the electronic nose endoscopic image into a nasopharyngeal carcinoma diagnosis model based on the image segmentation convolutional neural network, obtaining a malignant tumor region marked by the diagnosis model, and carrying out the positioning segmentation of the nasopharyngeal carcinoma. The diagnosis system can judge the captured image in real time, mark the malignant tumor part in the nasopharyngeal carcinoma image, export the diagnosis result, intuitively judge whether the target lesion is the malignant tumor tissue or not and determine the boundary range of the malignant tumor lesion according to the malignant tumor lesion as long as the lens is focused on the suspicious lesion tissue in the nasopharyngeal cavity. Suspicious diseased regions are quickly selected for biopsy, so that the accuracy of nasopharyngeal carcinoma detection under a nasal endoscope is effectively improved, and the detection rate of biopsy is increased.

Description

technical field [0001] The invention belongs to the field of medical diagnosis, and in particular relates to a method and system for diagnosing the nature and scope of nasopharyngeal carcinoma by using artificial intelligence to identify a nasal endoscopic image and capture a malignant tumor target region (ROI). Background technique [0002] Nasopharyngeal carcinoma (NPC) refers to malignant tumors that occur on the top and side walls of the nasopharyngeal cavity, and is one of the high-risk malignant tumors. The current electronic nasal endoscopy uses ordinary white light imaging (WLI) as the lighting device, which has certain limitations in clarity and contrast, and it is easy to miss the diagnosis of superficial early cancer and precancerous lesions that occur on the mucosal surface. Narrow band imaging (narrow band imaging, NBI) technology uses a narrow-spectrum filter to remove the red light in the ordinary endoscope, and only releases two wavelengths of light with a ce...

AI Technical Summary

Problems solved by technology

[0005] However, the current artificial intelligence nasal endoscopy-assisted diagnosis system is a binary classification diagnosis model. The system judges the nasopharynx image as "malignant tumor" or "non-malignant tumor" by calculating the probability of nasal endoscopic images. As a result, it is impossible to make an exact judgment on the exact location of the malignant tumor in the image

Images