| The uneven distribution of medical resources,the large shortage of doctors,and the continuous increase of clinical work intensity are the prominent difficulties faced by my country’s medical and health undertakings.The use of computer-aided diagnosis technology to promote the development of my country’s medical informatization has become a realistic demand.With the development of information technology and the popularization of smart devices,augmented reality technology has become an important driving force for the development of the industry.Head-mounted interactive device is an important carrier of augmented reality technology,which can provide people with immersive experience,and its application in the medical field has become a hot issue in current research.This dissertation combines medical image-assisted diagnosis and augmented reality technology to provide users with a natural and harmonious interactive process for auxiliary diagnosis.This dissertation firstly studies the multi-label semantic annotation algorithm of CT(Computed Tomography)images of multiple parts of the human body,and then combines the augmented reality technology to design and implement a CT image lesion detection and semantic annotation system.The main work is as follows:(1)A method for semantic labeling of lesions in CT images is proposed.Most of the existing semantic labeling algorithms are aimed at a specific part of the human body or a specific type of lesion.In this dissertation,a general lesion dataset is used to train the model,which can realize the semantic labeling and retrieval of lesions in 8 parts of the human body.Mechanism and residual module are used to improve the ability of the model to extract the features of lesions.Finally,the effectiveness of the method proposed in this dissertation is proved by experiments.(2)A CT image lesion detection and annotation system based on Holo Lens is designed and implemented.In order to solve the problem of limited computing performance and storage space of Holo Lens,the system is designed based on the clientserver structure.The server is mainly responsible for the lesion detection,labeling and retrieval of CT images.The head-mounted augmented reality interactive device Holo Lens is used as the client.Responsible for CT image acquisition,result presentation and multisensory interaction.The results of system operation show that the system achieves the purpose of CT image aided diagnosis. |
