Research On Disease Classification And Recognition Method Based On Children's Facial Images

As we all know,Chinese medicine pays attention to "Wang Wen Wen Qi",in which "Wang" refers to the observation of the patient's facial features,so as to initially judge the condition.But this requires a certain amount of experience.In recent years,the research on objectiveization and computer-aided diagnosis of "look" in traditional Chinese medicine has also become a research focus of computer-aided diagnosis with the development of image acquisition and computer big data technology.In order to improve the diagnosis rate and cure rate,there are many complicated children's diseases.At present,traditional Chinese medicine "Wang" method is also used for auxiliary diagnosis,that is,auxiliary recognition and diagnosis based on facial images,especially for children's growth and development diseases,such as special Nasal syndrome,hypothyroidism and other diseases,because children are young,can not express their symptoms clearly and delay treatment.This paper proposes a disease classification and recognition method based on children's facial images in response to the above-mentioned computer-aided diagnosis of children's developmental diseases.This has a certain reference value to solve the problem of medical treatment for children in remote areas.The main work and innovations of this article are as follows:1.Collect facial images about children's growth and development,establish a database,and complete image preprocessing.2.First,a convolutional neural network(CNN)combined with a support vector machine(Support Vector Machine,SVM)hybrid method(CNN+SVM)is used to treat Turner syndrome,precocious puberty,hypothyroidism,and breast development.Six categories of poor,short,and lack of growth hormone.Recognize the growth and development diseases of children,select the three categories that are strongly correlated with facial image features,and further subdivide,and finally select the Turner syndrome is the most suitable for using the face The image was used for auxiliary diagnosis,and an experimental study was carried out on the disease using the karyotype classification method.The research shows that the method proposed in this article has a good recognition effect on Turner syndrome,hypothyroidism,and growth hormone deficiency.Among them,Turner syndrome is best suited for assisted diagnosis of facial images.3.The facial image discrimination method of Turner karyotype was explored,and the improved VGG16 network and hybrid framework were compared and analyzed.Experiments proved that different karyotypes are distinguishable.