Applying surface-based morphometry to study ventricular abnormalities of cognitively unimpaired subjects prior to clinically significant memory decline

Abstract

Highlights A completely automated surface-based ventricular morphometry system. Generate a whole connected 3D ventricular shape model. Test-retest the system in two independent CU subject cohorts. Subregional ventricular abnormalities prior to clinically memory decline.

Predicting Brain Amyloid using Multivariate Morphometry Statistics, Sparse Coding, and Correntropy: Validation in 1,101 Individuals from the ADNI and OASIS Databases

“…The MMS measure for brain structures performs well in clinical group comparisons (Dong et al, 2020b, 2019; Li et al, 2016; Shi et al, 2015, 2014b; Wang et al, 2013; Yao et al, 2018), and as we have shown, it has the potential to further be applied for individual Aβ classification. To achieve this goal, we need to solve the challenge that the MMS dimension is usually much larger than the number of subjects, i.e., the so-called high dimension, small sample size problem .…”

“…Second, this work only applies hippocampal MMS to predict Aβ positivity. In future work, we plan to introduce more AD risk factors (such as demographic information, genetic information and clinical assessments) (Ansart et al, 2020; Pekkala et al, 2020; Tosun et al, 2014), and more AD regions of interest (ROIs, e.g., ventricles, entorhinal cortex, temporal lobes) (Brier et al, 2016; Dong et al, 2020b; Foley et al, 2017) into our proposed framework; these additional features are expected to improve the Aβ positivity prediction.…”

“…This approach performs better for detecting clinically-relevant group differences, relative to other TBM-based methods including those using the Jacobian determinant, the largest and smallest eigenvalues of the surface metric and the pair of eigenvalues of the Jacobian matrix (Wang et al, 2011, 2010). Our recent studies (Dong et al, 2020b, 2019) show that MMS outperforms volume measures for detecting hippocampal and ventricular deformations in groups at high risk for AD at the preclinical stage. Our other related work (Wu et al, 2018) has studied hippocampal morphometry in cohorts consisting of Aβ positive AD patients (Aβ+ AD) and Aβ negative cognitively unimpaired subjects (Aβ-CU) using the MMS measure.…”

“…Pekkala et al (2020) use brain MRI measures (volumes of the cortical gray matter, hippocampus, accumbens, thalamus and putamen) to infer Aβ positivity in cognitively unimpaired (CU) subjects; they achieve a 0.70 area under the receiver operator curve (AUC) with their Disease State Index (DSI) algorithm. Although brain structural volumes are perhaps the most commonly used neuroimaging measures in AD research (Cacciaglia et al, 2018; Crivello et al, 2010; Reiter et al, 2017), surface-based subregional structure measures can offer advantages over volume measures as they contain more detailed and patient-specific shape information (Apostolova et al, 2010; Ching et al, 2020; Costafreda et al, 2011; Dong et al, 2020b, 2019; Morra et al, 2009; Qiu et al, 2009; Shen et al, 2009; Styner et al, 2004; Paul M Thompson et al, 2004; Younes et al, 2014).…”

Predicting Brain Amyloid Using Multivariate Morphometry Statistics, Sparse Coding, and Correntropy: Validation in 1,101 Individuals From the ADNI and OASIS Databases

“…Pekkala et al (2020) use brain MRI measures (volumes of the cortical gray matter, hippocampus, accumbens, thalamus, and putamen) to infer Aβ positivity in cognitively unimpaired (CU) subjects; they achieve a 0.70 area under the receiver operator curve (AUC) with their Disease State Index (DSI) algorithm. Although brain structural volumes are perhaps the most commonly used neuroimaging measures in AD research (Crivello et al, 2010;Reiter et al, 2017;Cacciaglia et al, 2018), surface-based subregional structure measures can offer advantages over volume measures as they contain more detailed and patient-specific shape information (Styner et al, 2004;Thompson et al, 2004;Morra et al, 2009;Qiu et al, 2009;Shen et al, 2009;Apostolova et al, 2010;Costafreda et al, 2011;Younes et al, 2014;Dong et al, 2019Dong et al, , 2020bChing et al, 2020).…”