Problem Statement

Tooth wear is a widespread issue within the population where teeth degrade over time due to abrasion, attrition, and erosion. According to the NHS, 77% of adults experience some degree of tooth wear. Due to various treatment approaches, dentists often struggle with uncertainty in providing diagnosis and prognosis.

Solution

The project proposes a solution to develop a clinical decision support system (CDSS) app for windows that can assist dentists in accurately providing optimal treatment plans. For this to be possible, the app must be compatible with the PLY/STL file format and mantain a database in which to store these models along with patient metadata which dentists input themself. The app will use deep learning algorithms to analyse the 3D tooth models along with patient data to provide suggestions which support dentists in the process of diagnosing and deciding prognosis.

Achievement & Impact

Our team is developing a machine learning (ML) based windows app that uses 3D teeth models to suggest treatment plans. This includes the tooth wear analysis of PLY tooth models using deep learning, along with patient metadata to predict optimal treatment plans. This is done in our app using Python through a variety of libraries such as PySide2 and Open3D for our frontend along with Pytorch and SQLite for the backend. Overall, our app achieves a 91% accuracy on predicting tooth wear and has the potential to improve the efficiency and accuracy of tooth wear assessment with larger data set, leading to better oral health outcomes for patients.