Identifying individuals with Alzheimer's disease‐like brains based on structural imaging in the Human Connectome Project Aging cohort

Abstract

Given the difficulty in factoring out typical age effects from subtle Alzheimer's disease (AD) effects on brain structure, identification of very early, as well as younger preclinical "at-risk" individuals has unique challenges. We examined whether agecorrection procedures could be used to better identify individuals at very early potential risk from adults who did not have any existing cognitive diagnosis. First, we obtained cross-sectional age effects for each structural feature using data from a selected portion of the Human Connectome Project Aging (HCP-A) cohort. After age detrending, we weighted AD structural deterioration with patterns quantified from data of the Alzheimer's Disease Neuroimaging Initiative. Support vector machine was then used to classify individuals with brains that most resembled atrophy in AD across the entire HCP-A sample. Additionally, we iteratively adjusted the pipeline by removing individuals classified as AD-like from the HCP-A cohort to minimize atypical brain structural contributions to the age detrending. The classifier had a mean crossvalidation accuracy of 94.0% for AD recognition. It also could identify mild cognitive impairment with more severe AD-specific biomarkers and worse cognition. In an independent HCP-A cohort, 8.8% were identified as AD-like, and they trended toward worse cognition. An "AD risk" score derived from the machine learning models also significantly correlated with cognition. This work provides a proof of Complete listing of ADNI investigators can be found in the Data availability statement

Multiscale structural mapping of Alzheimer’s disease neurodegeneration

“…This work demonstrates the possibility to use this technique, which requires only a single T1-weighted MRI, in diagnostic support clinically, as well as to screen individuals who are likely to be a candidate for more intensive biomarker assessment. Future work will apply the MSSM procedures in individuals with earlier impairment and/or pathology stages, MCI with conversion to other types of dementia, and preclinical at-risk individuals ( Li et al, 2021 ) to determine the benefits of these novel features.…”

Adjustment for the Age- and Gender-Related Metabolic Changes Improves the Differential Diagnosis of Parkinsonism

“…It is believed that age-related changes may have a greater impact on imagebased diagnosis in patients outside the common onset ages because changes associated with later-life disease are subtle at very young patients, and age-related brain decline is severe in very old patients, which is leading to overestimation of those disease-free areas (Dukart et al 2011). Hence, age correction methods have been proposed for both MRI (Dukart et al 2011;Li et al 2021) and [ 18 F]FDG-based (Jiang et al 2018) diagnosis of dementia. The onset age for parkinsonism is generally around 55-65 years old, and some early onset cases may show symptoms in the early 20s, while some patients show symptoms at very old ages (i.e., late 80s).…”

Classifying Alzheimer's disease and frontotemporal dementia using machine learning with cross‐sectional and longitudinal magnetic resonance imaging data

“…In addition, in a previous study including both diseases, we showed that distinct brain atrophy patterns could potentially help in differentiating AD and FTD (Falgàs et al, 2020 ). More recently, measures derived from MRI have been used within ML algorithms to differentiate these diseases (Bron et al, 2017 ; Li et al, 2021 ; Möller et al, 2016 ; Penny et al, 2007 ). These approaches have shown good results.…”