Prototyping Digital Tongue Diagnosis System On Roborealm And Raspberry-PI

Abstract

In Oriental Medicine, the examination of tongue body contributes to rich information in diagnosing some diseases. The practitioners use tongue body colour and other relevant features to predict the patient’s conditions and disease-oriented state. To reduce the ambiguity and other limitation caused by environment, an automatic tongue acquisition system has been introduced recently to assist the medical practitioner in daily clinical practices. However, most of these acquisition systems are not equipped with the intelligent decision support system to diagnose certain diseases using informative tongue bio-markers. In this paper, automatic tongue diagnosis prototype on Raspberry-pi is proposed to reduce human labour and to promote portability. To integrate with Internet of Things (IOT) technology, tongue image processing analysis including segmentation and colour classification techniques using k-means and Support Vector Machine (SVM) is executed on Raspberry-Pi that is equipped with High Definition (HD) camera via MATLAB Coder platform. This high definition camera will capture the real time image of a patient’s tongue to be segmented and analysed further. It is expected to provide health information that can be used as a pre-diagnosis system based on Oriental Medicine perspectives.