Computer Aided Tongue Diagnosis System for Disease Prediction using Machine Learning Techniques

Abstract

newline newlineABSTRACT newlineHealth Informatics is an emerging field used to help the community to have a quality newlineof life. It is combination of the Medical Science and the Computer Engineering. newlineMedical practitioners are having huge amount of data of their patients that was collected newlineby through counselling and in the form medical reports. Computer Science techniques newlineprovides provision to store this data to do further analysis and examine to provide better newlinetreatment for their patients. Basically, Health Informatics is part of multidisciplinary newlinefields that combines medical field with computational approaches to support medical newlinepractitioners to serve the society better. newline newlineFrom ancient days, medical practitioners examine patients by using pulse, eyes, face newlineobservation, tongue etc. Tongue observation involves complexity as it carries lot of newlineinformation. Each area on the Tongue along with colour and coating provides different newlineobservations for the medical practitioner. It requires high experience to analyse the newlineinternal state of organs by observing the Tongue. The Tongue diagnosis is an important newlineway of monitoring human health status in Indian Ayurvedic Medicine(IAM), which newlinehelps to identify the different diseases of human through Tongue image analysis. The newlinemain is to study and analyse the Tongue through Tongue image to predict a disease newlineusing computer aided Machine Learning Techniques. newline newlineThe proposed model focused on Artificial Intelligence framework for Disease newlinePredication through Tongue Image Analysis (DPTIA). Initially, Fast Non-Local Mean newline(FNLM) filtering is applied on test image, which performs the different noise removal newlineoperation, colour enhancements and pre-processing operations. Here, pre-processing newlineoperation is performed to equalize the image resolutions in dataset. In addition, Grey newlineLevel Cooccurrence Matrix (GLCM) is also used to extract the texture features. Finally, newlineHybrid Extreme Learning Machine (HELM) classifier is used to classify the different newlinediseases from the extracted features. The DPTIA model is ca