Exploiting HPC resources for the 3D-time series analysis of caries lesion activity

Abstract

We present a research framework to analyze 3D-time series caries lesion activity based on collections of SkyScan® μ-CT images taken at different times during the dynamic caries process. Analyzing caries progression (or reversal) is data-driven and computationally demanding. It involves segmenting high-resolution μ-CT images, constructing 3D models suitable for interactive visualization, and analyzing 3D and 4D (3D + time) dental images. Our development exploits XSEDE's supercomputing, storage, and visualization resources to facilitate the knowledge discovery process. In this paper, we describe the required image processing algorithms and then discuss the parallelization of these methods to utilize XSEDE's high performance computing resources. We then present a workflow for visualization and analysis using ParaView. This workflow enables quantitative analysis as well as three-dimensional comparison of multiple temporal datasets from the longitudinal dental research studies. Such quantitative assessment and visualization can help us to understand and evaluate the underlying processes that arise from dental treatment, and therefore can have significant impact in the clinical decision-making process and caries diagnosis.