11 May 2020

Molar Segmentation Using Watershed, Top-Hat, and Surface Generation

Data, tutorial, and complete project for segmenting a CT-scanned human molar artifact. The tutorial and project cover advanced image segmentation and surface generation.


The "Data" folder contains three datasets. The file "human_molar_resampled_50um3.am" is a computed tomography scan of a human molar that was initially imaged with 15 um isotropic voxels. The original data were cropped and downsampled to 50 um isotropic voxels and saved in Avizo binary format. This file can also be read in Amira software. It is an 8-bit grayscale dataset. The other two files "Segmentation_DentinEnamel.am" and "Segmentation_PulpDentinEnamelCracks.am" are Avizo-format 8-bit label datasets that correspond with the downsampled CT scanned tooth.

The folder "AvizoProject" contains a reproducible Avizo project and associated data files. The "human_molar_segmentation-files" must be located in the same directory as the project file "human_molar_segmentation.hx" to be opened correctly. If Avizo's Spatial Units Management is turned on in the Avizo preferences, the data should load with correct voxel sizes (50 um^3).

The folder "Images" contains a few representative images of the dataset and the outputs of this workflow.

The folder "OtherOutputs" contains an Excel-readable .xml file with the output of the label analysis of the molar's label information. This directory also contains a .obj surface file that could be viewed in third-party software capable of rendering surfaces. Finally, the file "human_molar_resampled_50um3_tophat.labels.tif" is a 3D tiff stack created from the label data generated from this workflow. Most third party image analysis tools (e.g., FIJI) can handle this data structure, though the color information may be altered compared to viewing in Avizo.

Finally, the parent directory also contains a guided walkthrough in pdf format of the segmentation of the tooth. It is written for new Avizo users.

Data are courtesy of Dr. Timothy Ryan, Department of Anthropology, Pennsylvania State University