Monitoring method and system for endoscope diagnosis and treatment operation process

Abstract

The invention provides a monitoring method and system for an endoscope diagnosis and treatment operation process. The method comprises the steps of according to historical continuous frames of endoscopic images, detecting a rectangular area where an endoscopic visual field is located in each frame of endoscopic image, intercepting a rectangular area image and converting the rectangular area imageinto a grayscale image to form a sample set; labeling a label to which each sample image belongs; training a full-connection neural network by utilizing the sample set labeled with labels; for continuous k frames of grayscale images, if the grayscale images exceeding a preset proportion threshold belong to the body, judging that an endoscope enters the body, otherwise, judging that the endoscope is in vitro; receiving continuous k frames of endoscope images in real time from the start moment of endoscope diagnosis and treatment operation, inputting the images into the trained full-connection neural network, and outputting the position of the endoscope; if the endoscope does not enter the body and is still outside the body, y frames of endoscope images are moved backwards, wherein y is smaller than k; and taking the continuous k frames of endoscope images again and inputting into the trained full-connection neural network to judge the position of the endoscope again, wherein k is a positive integer larger than or equal to 3.

Description

Technical field [0001] The invention belongs to the field of endoscopic diagnosis and treatment image processing, and in particular relates to a monitoring method and system for an endoscopic diagnosis and treatment operation process. Background technique [0002] The statements in this section merely provide background information related to the present invention, and do not necessarily constitute prior art. [0003] Due to the large differences in operator behavior during digestive endoscopy, for example, insufficient inspection time, unstable inspection field, etc., the possibility of missing lesions is increased and the quality of endoscopy is reduced. However, the inventor found that traditional endoscope quality supervision methods consume a lot of labor and cost, and lack objective quantification methods, and cannot be carried out at all levels of endoscope centers. Summary of the invention [0004] In order to solve the above-mentioned problems, the first aspect of the pres...

AI Technical Summary

Problems solved by technology

However, the inventors found that the traditional method of endoscopy quality supervision consumes a lot of labor costs, and lacks objective quantification methods, so it cannot be carried out in endoscopy centers at all levels.

Images