A Comprehensive Segmentation, Registration, and Cancer Detection Scheme on 3 Tesla In Vivo Prostate DCE-MRI

  • Satish Viswanath
  • B. Nicolas Bloch
  • Elisabeth Genega
  • Neil Rofsky
  • Robert Lenkinski
  • Jonathan Chappelow
  • Robert Toth
  • Anant Madabhushi
Conference paper

DOI: 10.1007/978-3-540-85988-8_79

Part of the Lecture Notes in Computer Science book series (LNCS, volume 5241)
Cite this paper as:
Viswanath S. et al. (2008) A Comprehensive Segmentation, Registration, and Cancer Detection Scheme on 3 Tesla In Vivo Prostate DCE-MRI. In: Metaxas D., Axel L., Fichtinger G., Székely G. (eds) Medical Image Computing and Computer-Assisted Intervention – MICCAI 2008. MICCAI 2008. Lecture Notes in Computer Science, vol 5241. Springer, Berlin, Heidelberg

Abstract

Recently, high resolution 3 Tesla (T) Dynamic Contrast-Enhanced MRI (DCE-MRI) of the prostate has emerged as a promising modality for detecting prostate cancer (CaP). Computer-aided diagnosis (CAD) schemes for DCE-MRI data have thus far been primarily developed for breast cancer and typically involve model fitting of dynamic intensity changes as a function of contrast agent uptake by the lesion. Comparatively there is relatively little work in developing CAD schemes for prostate DCE-MRI. In this paper, we present a novel unsupervised detection scheme for CaP from 3 T DCE-MRI which comprises 3 distinct steps. First, a multi-attribute active shape model is used to automatically segment the prostate boundary from 3 T in vivo MR imagery. A robust multimodal registration scheme is then used to non-linearly align corresponding whole mount histological and DCE-MRI sections from prostatectomy specimens to determine the spatial extent of CaP. Non-linear dimensionality reduction schemes such as locally linear embedding (LLE) have been previously shown to be useful in projecting such high dimensional biomedical data into a lower dimensional subspace while preserving the non-linear geometry of the data manifold. DCE-MRI data is embedded via LLE and then classified via unsupervised consensus clustering to identify distinct classes. Quantitative evaluation on 21 histology-MRI slice pairs against registered CaP ground truth estimates yielded a maximum CaP detection accuracy of 77.20% while the popular three time point (3TP) scheme yielded an accuracy of 67.37%.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

Copyright information

© Springer-Verlag Berlin Heidelberg 2008

Authors and Affiliations

  • Satish Viswanath
    • 1
  • B. Nicolas Bloch
    • 2
  • Elisabeth Genega
    • 2
  • Neil Rofsky
    • 2
  • Robert Lenkinski
    • 2
  • Jonathan Chappelow
    • 1
  • Robert Toth
    • 1
  • Anant Madabhushi
    • 1
  1. 1.Department of Biomedical EngineeringRutgers UniversityUSA
  2. 2.Department of Radiology, Beth Israel Deaconess Medical Center USA

Personalised recommendations