Quantitative assessment of field strength, total intracranial volume, sex, and age effects on the goodness of harmonization for volumetric analysis on the ADNI database

Ma, Da; Popuri, Karteek; Bhalla, Mahadev; Sangha, Oshin; Lu, Donghuan; Cao, Jiguo; Jacova, Claudia; Wang, Lei; Beg, Mirza Faisal; Initiative, Alzheimer’s Disease Neuroimaging

doi:10.1002/hbm.24463

Cited by 41 publications

(47 citation statements)

References 104 publications

(169 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Though much of the early studies on how brain areas change in size over the lifespan utilized crosssectional data, there has been an increase in the number of studies using longitudinal data collection. One caution on relying on estimates based on cross-sectional data alone is the variability added due to differences in scanner properties and the sex of the subjects [26,27]. Though the current analysis focused on cross-sectional data, there has been good consistency found in the estimated rate of change when cross-sectional data was compared to longitudinal data in older adults ( [3,14] though see [28]).…”

Section: Discussionmentioning

confidence: 91%

Validation of Critical Ages in Regional Adult Brain Maturation

Nichols

2020

Front. Appl. Math. Stat.

View full text Add to dashboard Cite

Some aspects of the brain change volume at a fairly constant rate across the life span, i.e., linear maturation rate with respect to age, whereas other aspects of the brain change volume at different rates for younger vs. older age ranges, i.e., non-linear changes. Various forms of non-linear maturation likely reflect different biological mechanisms such that theoretical distinctions between maturation patterns ought to be considered. Simulated data with known maturation patterns and a single critical age characterizing a qualitative change in maturation were used to establish the validity of a non-parametric fitting method, the smoothing spline, combined with processing steps for determining the form of the pattern and the associated critical age. Multiple classes of models were assessed, including model-free, bottom up approaches. Three categories of maturation patterns were explored: U-shaped, with a change in direction; sigmoidal, with an isolated period of change; accelerating, with changes in amplitude but not direction. As noise is a limiting factor in curve fitting, smoothing splines were fit to data with idealized low and realistic noise levels. The smoothing spline was shown to contain the relevant information to extract the critical ages of all maturation patterns in the form of derivative zero points, but the previously proposed method of using third derivative zero points worked only for the accelerating category. Therefore, an additional classification step was included to first determine the category of maturation pattern. Classification accuracy and identification of the calculated critical age within 5 years of the actual critical age was found to be perfect for low noise and high for realistic noise levels. To demonstrate the applicability of the method, a reevaluation of published biological data previously analyzed using third derivative zero points to determine critical ages was carried out for 17 aspects of MRI scans from 1,100 subjects. For a majority of non-linear areas, new critical ages were identified. Further modifications to the analysis procedure could include a wider set of maturation patterns and the inclusion of multiple critical ages to help determine distinctions between brain areas in the timing of developmental or degenerative events that influence their volume.

show abstract

Section: Discussionmentioning

confidence: 91%

Validation of Critical Ages in Regional Adult Brain Maturation

Nichols

2020

Front. Appl. Math. Stat.

View full text Add to dashboard Cite

show abstract

“…2.8 | Assess the effect of demographic-and scanner-related variables on MRDATS It has been shown that both demographic-related (sex and age) as well as the scanner-related (field strength) variables might influence the outcomes of an analysis involving MRI-based volume features (Ma, Popuri, et al, 2019). Therefore, we explored the effect of demographic-related (sex and age) as well as the scanner-related variables (field strength) on…”

Section: Comparison Between Mrdats With Csf Biomarkermentioning

confidence: 99%

Using machine learning to quantify structural MRI neurodegeneration patterns of Alzheimer's disease into dementia score: Independent validation on 8,834 images from ADNI, AIBL, OASIS, and MIRIAD databases

Popuri

Wang

et al. 2020

Human Brain Mapping

Self Cite

View full text Add to dashboard Cite

Biomarkers for dementia of Alzheimer's type (DAT) are sought to facilitate accurate prediction of the disease onset, ideally predating the onset of cognitive deterioration. T1‐weighted magnetic resonance imaging (MRI) is a commonly used neuroimaging modality for measuring brain structure in vivo, potentially providing information enabling the design of biomarkers for DAT. We propose a novel biomarker using structural MRI volume‐based features to compute a similarity score for the individual's structural patterns relative to those observed in the DAT group. We employed ensemble‐learning framework that combines structural features in most discriminative ROIs to create an aggregate measure of neurodegeneration in the brain. This classifier is trained on 423 stable normal control (NC) and 330 DAT subjects, where clinical diagnosis is likely to have the highest certainty. Independent validation on 8,834 unseen images from ADNI, AIBL, OASIS, and MIRIAD Alzheimer's disease (AD) databases showed promising potential to predict the development of DAT depending on the time‐to‐conversion (TTC). Classification performance on stable versus progressive mild cognitive impairment (MCI) groups achieved an AUC of 0.81 for TTC of 6 months and 0.73 for TTC of up to 7 years, achieving state‐of‐the‐art results. The output score, indicating similarity to patterns seen in DAT, provides an intuitive measure of how closely the individual's brain features resemble the DAT group. This score can be used for assessing the presence of AD structural atrophy patterns in normal aging and MCI stages, as well as monitoring the progression of the individual's brain along with the disease course.

show abstract

“…Finally, in the feature extraction and normalization step, the volume of each patch was extracted as a primary feature for disease classification. The w-score, which represents the standardized residual of the chosen features, was computed to remove the effect of covariates such as the field of strength (1.5T or 3T), scanner type, scanning site, age, sex, and the size of the intra-cranial vault (ICV) of each individual (Ma et al, 2018 andPopuri et al, 2020). The normalized features as represented by the w-scores were input into the classifier.…”

Section: Multi-level Multi-type Feature Extractionmentioning

confidence: 99%

“…The locally-clustered cortical patches were then propagated back to each of the target space following the backward deformation field that was derived during the LDDMM non-rigid registration step (Beg et al, 2005). The average thickness of the mantle within each patch was computed as features followed by the w-score normalization (Ma et al, 2018 andPopuri et al, 2020) to remove the confounding effect of covariates.…”

Section: Multi-level Multi-type Feature Extractionmentioning

confidence: 99%

Differential Diagnosis of Frontotemporal Dementia, Alzheimer's Disease, and Normal Aging Using a Multi-Scale Multi-Type Feature Generative Adversarial Deep Neural Network on Structural Magnetic Resonance Images

Popuri

et al. 2020

Front. Neurosci.

Self Cite

View full text Add to dashboard Cite

Methods: Alzheimer's disease and Frontotemporal dementia are the first and third most common forms of dementia. Due to their similar clinical symptoms, they are easily misdiagnosed as each other even with sophisticated clinical guidelines. For disease-specific intervention and treatment, it is essential to develop a computer-aided system to improve the accuracy of their differential diagnosis. Recent advances in deep learning have delivered some of the best performance for medical image recognition tasks. However, its application to the differential diagnosis of AD and FTD pathology has not been explored. Approach: In this study, we proposed a novel deep learning based framework to distinguish between brain images of normal aging individuals and subjects with AD and FTD. Specifically, we combined the multi-scale and multi-type MRI-base image features with Generative Adversarial Network data augmentation technique to improve the differential diagnosis accuracy. Results: Each of the multi-scale, multitype, and data augmentation methods improved the ability for differential diagnosis for both AD and FTD. A 10-fold cross validation experiment performed on a large sample of 1,954 images using the proposed framework achieved a high overall accuracy of 88.28%. Conclusions: The salient contributions of this study are threefold: (1) our experiments demonstrate that the combination of multiple structural features extracted at different scales with our proposed deep neural network yields superior performance than Ma et al. FTD AD Differential Diagnosis individual features; (2) we show that the use of Generative Adversarial Network for data augmentation could further improve the discriminant ability of the network regarding challenging tasks such as differentiating dementia sub-types; (3) and finally, we show that ensemble classifier strategy could make the network more robust and stable.

show abstract

Quantitative assessment of field strength, total intracranial volume, sex, and age effects on the goodness of harmonization for volumetric analysis on the ADNI database

Cited by 41 publications

References 104 publications

Validation of Critical Ages in Regional Adult Brain Maturation

Validation of Critical Ages in Regional Adult Brain Maturation

Using machine learning to quantify structural MRI neurodegeneration patterns of Alzheimer's disease into dementia score: Independent validation on 8,834 images from ADNI, AIBL, OASIS, and MIRIAD databases

Differential Diagnosis of Frontotemporal Dementia, Alzheimer's Disease, and Normal Aging Using a Multi-Scale Multi-Type Feature Generative Adversarial Deep Neural Network on Structural Magnetic Resonance Images

Contact Info

Product

Resources

About