JoWUA

Abstract

Here, we discuss some machine-learning applications that support saliva sensing for real-time health monitoring. Saliva has been identified as a promising non-invasive diagnostic gateway for various types of biomarkers, which is readily available in numerous pathological conditions. Data analysis, pattern discovery, and prediction and classification tasks are some of the important fields in which machine learning techniques have had the most success. For instance, in the context of saliva sensing, machine learning algorithms can aid in performing efficient analysis of the multi-dimensional nature of saliva data and complex elements associated with it, enabling the detection of the slightest variation in biomarkers that may be indicative of certain health conditions. Yet, the most potential in salivary sensing is likely to be in the early identification and monitoring of diseases such as diabetes, cancer, and infectious diseases by machine learning models. Say the algorithms are trained on saliva samples, for example. In this case, they would learn to recognize patterns of saliva that correspond with specific health issues and accurately diagnose them. Another idea is to build personal systems that monitor your health continuously and help screen for potential problems.