A visual analytics system for breast tumor evaluation.

OBJECTIVE: To develop a system for the interactive exploration and examination of histologically derived data that is associated with breast tumors and may be used to evaluate the histologic grade of the tumor. STUDY DESIGN: The system integrates pathologist-generated prognostic data with 2-dimensional (2-D) image analysis data, 2-D digital tissue cross-sections and annotations, 3-dimensional (3-D) tumor reconstructions and volumetric analysis, 3D spatial tumor display and recorded prognostic information from available cases in the Drexel University College of Medicine tumor databank. The system consists of 3 components: (1) a user interface for applying 2-D image processing, segmentation and annotation to a digitized histology slide, (2) a distance field interpolation method for contour-based 3D reconstruction of breast tumors and volumetric model analysis routines and (3) a Web-based database management interface for interactive data browsing and searching and multimodality visualization. RESULTS: The system has been implemented and deployed with data from 36 breast cancer cases, 7 of which have been reconstructed in 3-D. CONCLUSION: Interactive visual analytics technology may be used to create an effective breast tumor evaluation system.

Contour-Based Surface Reconstruction using MPU Implicit Models.

This paper presents a technique for creating a smooth, closed surface from a set of 2D contours, which have been extracted from a 3D scan. The technique interprets the pixels that make up the contours as points in ℝ(3) and employs Multi-level Partition of Unity (MPU) implicit models to create a surface that approximately fits to the 3D points. Since MPU implicit models additionally require surface normal information at each point, an algorithm that estimates normals from the contour data is also described. Contour data frequently contains noise from the scanning and delineation process. MPU implicit models provide a superior approach to the problem of contour-based surface reconstruction, especially in the presence of noise, because they are based on adaptive implicit functions that locally approximate the points within a controllable error bound. We demonstrate the effectiveness of our technique with a number of example datasets, providing images and error statistics generated from our results.