CSE Colloquium Series

3D General Lesion Segmentation in CT

Marie-Pierre Jolly

Senior Member, Technical Staff
Siemens Corporate Research, Princeton

Friday, March 14, 2008
Talk: 3:00 pm - 4:00 pm
3105 Engineering

Host: Anil Jain

Abstract

This talk describes a general purpose algorithm to segment any kind of esions in computer tomography images. The algorithm expects a click or a stroke inside the lesion from the user and learns gray level properties about the lesion on the fly. It then uses a random walker general purpose segmentation algorithm to combine multiple 2D segmentation results and produce the final 3D segmentation of the lesion. Quantitative evaluation on 293 lesions demonstrate that the method is ready for clinical use. We will also demonstrate the editing capabilities of the algorithm.