Automatic Cephalometric Analysis using Machine Learning

Abstract

Orthodontics is a specialized dental profession that specializes in diagnosing, preventing, and correcting teeth and jawbone, as well as biting patterns. The irregular shape of the teeth and jaws is very common. About 50% of the population of the developed world, according to the American Association of Orthodontics, has malocclusions heavy enough to benefit from orthopedic treatment. Cephalometries is often used by dentists as a tool for diagnosis and treatment planning and evaluation. Cephalometries aids in orthodontic diagnostics by empowering the study of skeletal structures, teeth, and soft tissues of the craniofacial region. Cephalometric analysis has many applications, including diagnostics, the definition of face measurement patterns, planning of orthodontic and orthognathic treatments, monitoring changes due to ageing or treatment and prediction of orthodontic and orthognathic treatment outcomes. Machine-based software is the only solution to reduce the dentist's work of planning and evaluation. Although some software are available for cephalometric analysis but, they are expensive and not easy to use as it requires heavy hardware-based tools such as laser guns and cephalostats. The proposed model uses a decision tree to develop a diagnostic program based on the data of previous patients assigned to it. This model is implemented using python language libraries such as Tkinter, Opencv, Sklearn, PIL and Pandas.