Block-based statistics for robust non-parametric morphometry

Geng Chen, Pei Zhang, Ke Li, Chong Yaw Wee, Yafeng Wu, Dinggang Shen, Pew Thian Yap

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

3 Scopus citations

Abstract

Automated algorithms designed for comparison of medical images are generally dependent on a sufficiently large dataset and highly accurate registration as they implicitly assume that the comparison is being made across a set of images with locally matching structures. However, very often sample size is limited and registration methods are not perfect and may be prone to errors due to noise, artifacts, and complex variations of brain topology. In this paper, we propose a novel statistical group comparison algorithm, called block-based statistics (BBS), which reformulates the conventional comparison framework from a non-local means perspective in order to learn what the statistics would have been, given perfect correspondence. Through this formulation, BBS (1) explicitly considers image registration errors to reduce reliance on high-quality registrations, (2) increases the number of samples for statistical estimation by collapsing measurements from similar signal distributions, and (3) diminishes the need for large image sets. BBS is based on permutation test and hence no assumption, such as Gaussianity, is imposed on the distribution. Experimental results indicate that BBS yields markedly improved lesion detection accuracy especially with limited sample size, is more robust to sample imbalance, and converges faster to results expected for large sample size.

Original languageEnglish
Title of host publicationPatch-Based Techniques in Medical Imaging - First st International Workshop, Patch-MI 2015 Held in Conjunction with MICCAI 2015, Revised Selected Papers
EditorsPierrick Coupé, Brent Munsell, Guorong Wu, Yiqiang Zhan, Daniel Rueckert
PublisherSpringer Verlag
Pages62-70
Number of pages9
ISBN (Print)9783319281933
DOIs
StatePublished - 2015
Event1st International Workshop on Patch-Based Techniques in Medical Imaging, Patch-MI 2015 - Munich, Germany
Duration: 9 Oct 20159 Oct 2015

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume9467
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference1st International Workshop on Patch-Based Techniques in Medical Imaging, Patch-MI 2015
Country/TerritoryGermany
CityMunich
Period9/10/159/10/15

Fingerprint

Dive into the research topics of 'Block-based statistics for robust non-parametric morphometry'. Together they form a unique fingerprint.

Cite this