Improving the SIENA performance using BEaST brain extraction

Kunio Nakamura,Douglas L Arnold,Simon F Eskildsen,D Louis Collins,Sridar Narayanan,Dzung Pham

doi:10.1371/journal.pone.0196945

Abstract

We present an improved image analysis pipeline to detect the percent brain volume change (PBVC) using SIENA (Structural Image Evaluation, using Normalization, of Atrophy) in populations with Alzheimer’s dementia. Our proposed approach uses the improved brain extraction mask from BEaST (Brain Extraction based on nonlocal Segmentation Technique) instead of the conventional BET (Brain Extraction Tool) for SIENA. We compared four varying options of BET as well as BEaST and applied these five methods to analyze scan-rescan MRIs in ADNI from 332 subjects, longitudinal ADNI MRIs from the same 332 subjects, their repeat scans over time, and OASIS longitudinal MRIs from 123 subjects. The results showed that BEaST brain masks were consistent in scan-rescan reproducibility. The cross-sectional scan-rescan error in the absolute percent brain volume difference measured by SIENA was smallest (p≤0.0187) with the proposed BEaST-SIENA. We evaluated the statistical power in terms of effect size, and the best performance was achieved with BEaST-SIENA (1.2789 for ADNI and 1.095 for OASIS). The absolute difference in PBVC between scan-dataset (volume change from baseline to year-1) and rescan-dataset (volume change from baseline repeat scan to year-1 repeat scan) was also the smallest with BEaST-SIENA compared to the BET-based SIENA and had the highest correlation when compared to the BET-based SIENA variants. In conclusion, our study shows that BEaST was robust in terms of reproducibility and consistency and that SIENA’s reproducibility and statistical power are improved in multiple datasets when used in combination with BEaST.

Highlights

Chronic brain atrophy is one of the pathologic hallmarks in neurological diseases such as multiple sclerosis and Alzheimer’s disease (AD)
The major advantages of the SIENA method are 1) skullconstrained registration, which corrects for incorrect pixel sizes [1] and reduces distortion artifacts [3], 2) halfway-space transformation whereby the images are interpolated, 3) measurement of edge shift using the first derivatives of the intensity profiles at the brain edge for sub-voxel accuracy, and 4) self-calibration to determine the ratio between surface area and volume to measure the percent volume change
Our previous work has shown that the performance of brain extraction in BET/FSL [5] worsened when applied to Alzheimer’s Disease Neuroimaging Initiative (ADNI) data compared to data from young normal subjects in the ICBM dataset in terms of Dice similarity coefficient (0.944 and 0.975), false positive rate (3.81 and 1.28), and false negative rate (2.71 and 0.45) [6]

Summary

Introduction

Chronic brain atrophy is one of the pathologic hallmarks in neurological diseases such as multiple sclerosis and Alzheimer’s disease (AD). The computations were made on the supercomputer Guillimin from McGill University, managed by Calcul Quebec and Compute Canada The operation of this supercomputer is funded by the Canada Foundation for Innovation (CFI), NanoQuebec, RMGA, the Fonds de recherche du Quebec - Nature et technologies (FRQ-NT) and a FRSQ-Pfizer Innovation grant. The Canadian Institutes of Health Research is providing funds to support ADNI clinical site in Canada. The use of SIENA in AD studies may bridge these neurological conditions (such as multiple sclerosis and Alzheimer’s disease) and allow metaanalysis to compare or merge such datasets in the future. We investigated the effect on SIENA measurements when applied in conjunction with BEaST-extracted brain masks. We 1) applied SIENA using various brain extraction approaches including BET and BEaST; 2) compared the extracted brain masks by measuring the overlaps between subjects and consistency within subjects; 3) compared SIENA scan-rescan reproducibility, and 4) compared the statistical power in terms of effect size and required sample size to detect significant changes using SIENA

Methods

Results

Conclusion