A Quantatiitive Study of Texture Features across Different Window Sizes in Prostate T2-weighted MRI

Andrik Rampun, Liping Wang, Paul Malcolm, Reyer Zwiggelaar

Research output: Contribution to journalArticlepeer-review

3 Citations (SciVal)
176 Downloads (Pure)

Abstract

This study aims to investigate the effects of window size on the performance of prostate cancer CAD and to identify discriminant texture descriptors in prostate T2-W MRI. For this purpose we extracted 215 texture features from 418 T2-W MRI images and extracted them using 9 different window sizes (3 × 3 to 19 × 19). The Bayesian Network and Random Forest classifiers were employed to perform the classification. Experimental results suggest that using window size of 9 × 9 and 11 × 11 produced Az > 89%. Also, this study suggests a set of best texture features based on our experimental results.
Original languageEnglish
Pages (from-to)74-79
Number of pages6
JournalProcedia Computer Science
Volume90
DOIs
Publication statusPublished - 06 Jul 2016
EventMedical Imaging Understanding and Analysis - Loughborough University, Loughborough, United Kingdom of Great Britain and Northern Ireland
Duration: 06 Jul 201608 Jul 2016
Conference number: 20

Keywords

  • texture descriptors
  • window size
  • computer aided diagnosis
  • T2-" MRI
  • prostate cancer

Fingerprint

Dive into the research topics of 'A Quantatiitive Study of Texture Features across Different Window Sizes in Prostate T2-weighted MRI'. Together they form a unique fingerprint.

Cite this