0 99

Cited 0 times in

Machine Learning for the Prediction of Amyloid Positivity in Amnestic Mild Cognitive Impairment

 Sung Hoon Kang  ;  Bo Kyoung Cheon  ;  Ji-Sun Kim  ;  Hyemin Jang  ;  Hee Jin Kim  ;  Kyung Won Park  ;  Young Noh  ;  Jin San Lee  ;  Byoung Seok Ye  ;  Duk L Na  ;  Hyejoo Lee  ;  Sang Won Seo 
 JOURNAL OF ALZHEIMERS DISEASE, Vol.80(1) : 143-157, 2021-03 
Journal Title
Issue Date
Aged ; Aged, 80 and over ; Algorithms ; Amyloid beta-Peptides / blood* ; Apolipoproteins E / genetics ; Area Under Curve ; Cerebral Cortex / diagnostic imaging ; Cognitive Dysfunction / diagnosis* ; Cognitive Dysfunction / diagnostic imaging ; Cognitive Dysfunction / psychology ; Cohort Studies ; Disease Progression ; Female ; Hippocampus / diagnostic imaging ; Humans ; Machine Learning* ; Magnetic Resonance Imaging ; Male ; Mental Recall ; Middle Aged ; Neuropsychological Tests ; Positron-Emission Tomography ; Predictive Value of Tests ; Reproducibility of Results ; Risk Factors
Aβ PET ; Aβ positivity ; amnestic mild cognitive impairment ; machine learning ; magnetic resonance imaging features ; neuropsychological tests ; prediction model
Background: Amyloid-β (Aβ) evaluation in amnestic mild cognitive impairment (aMCI) patients is important for predicting conversion to Alzheimer's disease. However, Aβ evaluation through Aβ positron emission tomography (PET) is limited due to high cost and safety issues.

Objective: We therefore aimed to develop and validate prediction models of Aβ positivity for aMCI using optimal interpretable machine learning (ML) approaches utilizing multimodal markers.

Methods: We recruited 529 aMCI patients from multiple centers who underwent Aβ PET. We trained ML algorithms using a training cohort (324 aMCI from Samsung medical center) with two-phase modelling: model 1 included age, gender, education, diabetes, hypertension, apolipoprotein E genotype, and neuropsychological test scores; model 2 included the same variables as model 1 with additional MRI features. We used four-fold cross-validation during the modelling and evaluated the models on an external validation cohort (187 aMCI from the other centers).

Results: Model 1 showed good accuracy (area under the receiver operating characteristic curve [AUROC] 0.837) in cross-validation, and fair accuracy (AUROC 0.765) in external validation. Model 2 led to improvement in the prediction performance with good accuracy (AUROC 0.892) in cross validation compared to model 1. Apolipoprotein E genotype, delayed recall task scores, and interaction between cortical thickness in the temporal region and hippocampal volume were the most important predictors of Aβ positivity.

Conclusion: Our results suggest that ML models are effective in predicting Aβ positivity at the individual level and could help the biomarker-guided diagnosis of prodromal AD.
Full Text
Appears in Collections:
1. College of Medicine (의과대학) > Dept. of Neurology (신경과학교실) > 1. Journal Papers
Yonsei Authors
Ye, Byoung Seok(예병석) ORCID logo https://orcid.org/0000-0003-0187-8440
사서에게 알리기


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.