ToothKit java software is intended to illustrate how data mining can be applied to practical analysis of 3D morphologies. The link between analysis and data mining is provided by probabilistic classifiers that can be "trained" to perform different tasks using various machine learning techniques. The current release is supplied with two types of classifiers: those trained to predict dietary class and those trained to predict morphological class. The first type is designed to make dietary predictions/reconstructions, while the second type is intended for automatic prototyping of teeth into broad dental types. In addition, both types are employed as search engines that can search your hard drive for morphologies (*.stl, *obj, *.dxf or *.wrl files) that are likely specialized in the same diet or that likely belong to the same dental type as a given sample.

In addition, ToothKit allows to convert *.stl, *obj, *.dxf and *.wrl polygon meshes into DEMs, visualize the original meshes and the DEMs, and to extract a multitude of topological features from the loaded morphologies.

The package is supplied with extensive help files that explain or reference the basic concepts behind the implementation.



Principal investigator: Jukka Jernvall
Concept and programming: Ilya Plyusnin
Morphology scanning and morphological expertise: Alistair Evans and Aleksis Karme


ToothKit is based on JavaView and WEKA Java libraries.


JavaView – 3D viewer Java application

WEKA – data mining software in Java


ToothKit is compatible with Java Runtime Environment 1.4 or higher. This software was tested and worked properly on:

  • JRE 1.5 on Microsoft Windows XP
  • JRE 1.4.2 on Mac OS 10.3.9


Click to download ToothKit1.1.

For additional information or to place comments contact ipljusni -at-