RSNA 2013 

Abstract Archives of the RSNA, 2013


Evaluation of WBAA with Registration-based Cube Propagation for Brain Atrophy Quantification

Scientific Informal (Poster) Presentations

Presented on December 5, 2013
Presented as part of LL-NRS-THB: Neuroradiology/Head and Neck - Thursday Posters and Exhibits (12:45pm - 1:15pm)


Martin Lillholm MSc, PhD, Presenter: Employee, Biomediq A/S Shareholder, Biomediq A/S
Akshay Pai, Abstract Co-Author: Nothing to Disclose
Lauge Sorensen, Abstract Co-Author: Nothing to Disclose
Mads Nielsen PhD, Abstract Co-Author: Stockholder, Biomediq A/S Research Grant, Nordic Bioscience A/S Research Grant, SYNARC Inc Research Grant, AstraZeneca PLC
Jon Sporring, Abstract Co-Author: Co-founder, DigiCorpus ApS Shareholder, DigiCorpus ApS
Sune Darkner, Abstract Co-Author: Nothing to Disclose
Erik B. Dam PhD, Abstract Co-Author: Shareholder, Biomediq A/S Employee, Biomediq A/S


Atrophy for the whole brain and sub-structures is becoming common as study outcome in clinical trials assessing the efficacy of potential treatments of diseases involving dementia. In this study, we evaluated the sensitivity to change  related to progression of Alzheimer's disease of a novel software framework, WBAA.


The recently defined Alzheimer's disease neuroimaging initiative (ADNI) standardized collection ("ADNI1:Annual 2 Yr 1.5T" at with 504 subjects (169 normals, 234 mild cognitively impaired, and 101 alzheimer) including baseline and 12-month 1.5T T1 magnetic resonance imaging (MRI) scans was used.  The MRIs were processed using longitudinal FreeSurfer and the whole brain atrophy application (WBAA 1.0 by Biomediq) that performs non-rigid registration followed by atrophy estimation quantification using cube propagation (CP). The WBAA was also evaluated with CP replaced by the common Jacobian integration (JI) method. Sensitivity to change was evaluated by atrophy differences between healthy and Alzheimer subjects quantified using Cohen's D and required study sample sizes.


As example, quantifications of the hippocampus atrophies estimated using WBAA were -1.3% and -0.6% for the Alzheimer and healthy subjects whereas the ventricle estimates were +9.3% and +4.1%, respectively. Corresponding Cohen's D for WBAA on these two regions were 1.1 and 1.0.  For whole-brain, hippocampus, ventricles, and medial temporal lobe, the WBAA Cohen's D were 0.7, 1.1, 1.0, and 1.3. The corresponding sample sizes were 173, 124, 113, and 87.  For WBAA with JI, Cohen's D were 0.5, 1.1, 1.0 and 1.2; with sample sizes 230, 139, 112, and 101.  For longitudinal FreeSurfer, Cohen's D were 0.7, 1.0, 1.0, and 1.3; with sample sizes 183, 152, 118, and 102.


The WBAA using CP for brain atrophy quantification provided sensitivity equal or superior to leading, competing methods. Specifically, the WBAA sample sizes were generally lower. 


Unlike longitudinal FreeSurfer, WBAA allows quantification of final atrophy estimates directly after each visit. Adding the matching/improved sensitivity, WBAA seems appropriate for clinical trials. 

Cite This Abstract

Lillholm, M, Pai, A, Sorensen, L, Nielsen, M, Sporring, J, Darkner, S, Dam, E, Evaluation of WBAA with Registration-based Cube Propagation for Brain Atrophy Quantification.  Radiological Society of North America 2013 Scientific Assembly and Annual Meeting, December 1 - December 6, 2013 ,Chicago IL.