Teeth Staining from Tobacco Consumption as an Indicator to Chronic Illness: A Data Analytics and Machine Learning Application

Tobacco consumption is a major public health concern because it is associated with a range of serious health problems, including lung cancer, heart disease, stroke, respiratory disease, and many others. Tobacco consumption is a leading cause of preventable death and disease worldwide, and it is estimated to be responsible for over 7 million deaths each year. The primary active ingredient in tobacco products is nicotine, which is highly addictive. People who use tobacco products can become physically dependent on nicotine, which can make it very difficult to quit using tobacco. Too much of nicotine consumption can cause the staining of the dental enamel. The aim of this study is to develop machine learning (ML) model to classify chronic diseases in person based on their dental staining. The study is divided into two phases. In the first phase involves the development of a smart system to detect and categorize the dental stain as high, average and low. This phase also involves the collection of demographic data and its analysis to derive valuable insights from it. The second phase involves training of the different ML models to recognize the hidden pattern in them. The parameters of the ML models are optimized by tuning and then the tuned ML models are employed for classifying the chronic illness. The Grid Searched Logistic Regression (GSLR) showed the best prediction out of all the ML models. The GSLR model showed training and testing accuracy of 64% and 60% respectively. The applicability of the model is ascertained by applying the model to the test data. The GSLR model is able to correctly classify 7 out of the 31 cases of no chronic diseases and 38 out of 44 cases of chronic diseases. Based on the applicability of the model, it can be concluded that the model is capable of detecting chronic diseases with dental enamel staining as an indicator.