Nonlinear registration of longitudinal images and measurement of change in regions of interest

Multimodal Imaging Laboratory, The University of California, San Diego, 9500 Gilman Drive, La Jolla, CA 92037, USA.
Medical image analysis (Impact Factor: 3.68). 02/2011; 15(4):489-97. DOI: 10.1016/
Source: PubMed

ABSTRACT We describe here a method, Quarc, for accurately quantifying structural changes in organs, based on serial MRI scans. The procedure can be used to measure deformations globally or in regions of interest (ROIs), including large-scale changes in the whole organ, and subtle changes in small-scale structures. We validate the method with model studies, and provide an illustrative analysis using the brain. We apply the method to the large, publicly available ADNI database of serial brain scans, and calculate Cohen's d effect sizes for several ROIs. Using publicly available derived-data, we directly compare effect sizes from Quarc with those from four existing methods that quantify cerebral structural change. Quarc produced a slightly improved, though not significantly different, whole brain effect size compared with the standard KN-BSI method, but in all other cases it produced significantly larger effect sizes.

1 Follower
  • Source
    • "Rates of whole brain and hippocampal atrophy from longitudinal magnetic resonane imaging (MRI) scans can aid in disease diagnosis and tracking of pathologic progression in neurodegenerative diseases and are increasingly used as outcome measures in trials of potentially disease-modifying therapies (Anderson et al., 2006; Frisoni et al., 2010; Holland et al., 2012; Sharma et al., 2010; Sluimer et al., 2010). Popular methods for brain atrophy measurement in longitudinal studies include Boundary Shift Integral (BSI) (Freeborough and Fox, 1997; Leung et al., 2010b, 2012), Structural Image Evaluation, using Normalization , of Atrophy (SIENA) (Smith et al., 2001), Quantitative Anatomical Regional Change (QUARC) (Holland and Dale, 2011), Tensor-Based Morphometry (TBM) (Hua et al., 2013), and FreeSurfer-longitudinal (FS) (Reuter et al., 2012). BSI and SIENA both use linear registration to align the baseline and repeat images and then track the shift of the brain boundary location, whereas QUARC and TBM both use nonlinear registrations to map between the baseline and repeat images and then measure volume change through analysis of the resulting deformation fields. "
    [Show abstract] [Hide abstract]
    ABSTRACT: Brain atrophy measured using structural MRI has been widely used as an imaging biomarker for disease diagnosis and tracking of pathological progression in neurodegenerative diseases. In this work, we present a generalised and extended formulation of the Boundary Shift Integral (gBSI) using probabilistic segmentations in order to estimate anatomical changes between 2 time points. This method adaptively estimates a non-binary XOR region-of-interest from probabilistic brain segmentations of the baseline and repeat scans, in order to better localise and capture the brain atrophy. We evaluate the proposed method by comparing the sample size requirements for a hypothetical clinical trial of Alzheimer’s disease to that needed for the current implementation of BSI as well as a fuzzy implementation of BSI. The gBSI method results in a modest, but reduced sample size, providing increased sensitivity to disease changes through the use of the probabilistic XOR region.
    Neurobiology of Aging 08/2014; 36. DOI:10.1016/j.neurobiolaging.2014.04.035 · 4.85 Impact Factor
  • Source
    • "A specific category of spatial deformations has spanned the attention of researchers in this field: diffeomorphisms (smooth deformation with a smooth inverse). These were largely used in different registration models (Allassonnière et al., 2005; Beg et al., 2005; Holland and Dale, 2011; Klein et al., 2009) and became a part of the classical deformable template theory – especially after the establishment of the Large Deformation Diffeomorphic Metric Mapping (LDDMM) – pioneered by Dupuis et al. (1998) and Trouvé (1995, 1998). The metamorphosis theory is built upon the LDDMM framework which is based on the idea of a diffeomorphic metric. "
    [Show abstract] [Hide abstract]
    ABSTRACT: We extend the image-to-image metamorphosis into constrained longitudi- nal metamorphosis. We apply it to estimate an evolution scenario, in patients with acute ischemic stroke, of both scattered and solitary ischemic lesions vis- ible on serial MR perfusion weighted imaging from acute to subacute stages. We then estimate a patient-specific residual map that enables us to capture the most relevant shape and intensity changes, continuously, as the lesion evolves from acute through subacute to chronic timepoints until merging into the final image. We detect areas with high residuals (ie. high dynamics) and identify areas that became part of the final T2-w lesion obtained at ≥ 1 month after stroke. This allows the investigation of the dynamic influence of perfusion values on the final lesion outcome as seen on T2-w imaging. The model provides detailed insights into stroke lesion dynamic evolution in space and time that will help identify factors that determine final outcome and identify targets for interventions to improve outcome.
    08/2014; 5:332–340. DOI:10.1016/j.nicl.2014.07.009
  • Source
    • "The analyses are performed on MRI scans from the ADNI-1 dataset. Using two follow-up time points, where available, in addition to the baseline scan allows us to estimate the presence of any longitudinal bias, or intransitivity, which has been a subject of controversy in recent ADNI studies (Hua et al., 2011, 2012; Thompson and Holland, 2011). To register the ventricles and compute radial thickness measures, we modify a recently proposed (Gutman et al., 2012) medial curve algorithm for longitudinal registration. "
    [Show abstract] [Hide abstract]
    ABSTRACT: We propose a new method to maximize biomarker efficiency for detecting anatomical change over time in serial MRI. Drug trials using neuroimaging become prohibitively costly if vast numbers of subjects must be assessed, so it is vital to develop efficient measures of brain change. A popular measure of efficiency is the minimal sample size (n80) needed to detect 25% change in a biomarker, with 95% confidence and 80% power. For multivariate measures of brain change, we can directly optimize n80 based on a Linear Discriminant Analysis (LDA). Here we use a supervised learning framework to optimize n80, offering two alternative solutions. With a new medial surface modeling method, we track 3D dynamic changes in the lateral ventricles in 2,065 ADNI scans. We apply our LDA-based weighting to the results. Our best average n80 - in two-fold nested cross-validation - is 104 MCI subjects (95% CI: [94,139]) for a 1-year drug trial, and 75 AD subjects [64,102]. This compares favorably with other MRI analysis methods. The standard "statistical ROI" approach applied to the same ventricular surfaces requires 165 MCI or 94 AD subjects. At 2 years, the best LDA measure needs only 67 MCI and 52 AD subjects, versus 119 MCI and 80 AD subjects for the stat-ROI method. Our surface-based measures are unbiased: they give no artifactual additive atrophy over three time points. Our results suggest that statistical weighting may boost efficiency of drug trials that use brain maps.
    NeuroImage 01/2013; 70. DOI:10.1016/j.neuroimage.2012.12.052 · 6.36 Impact Factor
Show more


1 Download
Available from