Optimized prostate biopsy via a statistical atlas of cancer spatial distribution.


Journal Article

A methodology is presented for constructing a statistical atlas of spatial distribution of prostate cancer from a large patient cohort, and it is used for optimizing needle biopsy. An adaptive-focus deformable model is used for the spatial normalization and registration of 100 prostate histological samples, which were provided by the Center for Prostate Disease Research of the US Department of Defense, resulting in a statistical atlas of spatial distribution of prostate cancer. Based on this atlas, a statistical predictive model was developed to optimize the needle biopsy sites, by maximizing the probability of detecting cancer. Experimental results using cross-validation show that the proposed method can detect cancer with a 99% success rate using seven needles, in these samples.

Full Text

Duke Authors

Cited Authors

  • Shen, D; Lao, Z; Zeng, J; Zhang, W; Sesterhenn, IA; Sun, L; Moul, JW; Herskovits, EH; Fichtinger, G; Davatzikos, C

Published Date

  • June 2004

Published In

Volume / Issue

  • 8 / 2

Start / End Page

  • 139 - 150

PubMed ID

  • 15063863

Pubmed Central ID

  • 15063863

International Standard Serial Number (ISSN)

  • 1361-8415

Digital Object Identifier (DOI)

  • 10.1016/j.media.2003.11.002


  • eng

Conference Location

  • Netherlands