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

Article type: Research Article

Authors: Kang, Sung Hoon^{a; b; c; 1} | Cheon, Bo Kyoung^{a; b; d; 1} | Kim, Ji-Sun^{a; b} | Jang, Hyemin^{a; b} | Kim, Hee Jin^{a; b} | Park, Kyung Won^e | Noh, Young^f | Lee, Jin San^g | Ye, Byoung Seok^h | Na, Duk L.^{a; b} | Lee, Hyejoo^{a; b; *} | Seo, Sang Won^{a; b; d; i; j; *}

Affiliations: [a] Departments of Neurology, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, Korea | [b] Neuroscience Center, Samsung Medical Center, Seoul, Korea | [c] Department of Neurology, Korea University Guro Hospital, Korea University College of Medicine, Seoul, Korea | [d] Department of Digital Health, SAIHST, Sungkyunkwan University, Seoul, Korea | [e] Department of Neurology, Dong-A University Medical Center, Dong-A University College of Medicine, Busan, Korea | [f] Department of Neurology, Gachon University Gil Medical Center, Incheon, Korea | [g] Department of Neurology, Kyung Hee University Hospital, Seoul, Korea | [h] Department of Neurology, Severance hospital, Yonsei University School of Medicine, Seoul, Korea | [i] Samsung Alzheimer Research Center and Center for Clinical Epidemiology Medical Center, Seoul, Korea | [j] Department of Intelligent Precision Healthcare Convergence, SAIHST, Sungkyunkwan University, Seoul, Korea

Correspondence: [*] Correspondence to: Sang Won Seo, MD, PhD, Department of Neurology, Sungkyunkwan University School of Medicine, Samsung Medical Center, 81 Irwon-ro, Gangnam-gu, Seoul 06351, Korea. Tel.:+82 2 3410 6147; Fax:+82 2 3410 0052; E-mail: sangwonseo@empal.com; sw72.seo@samsung.com. Hyejoo Lee, PhD, Department of Neurology, Sungkyunkwan University School of Medicine, Samsung Medical Center, 81 Irwon-ro, Gangnam-gu, Seoul 06351, Korea. Tel.:+82 2 2008 4335; Fax:+82 2 3410 0052; E-mail: hyejoo271@gmail.com.

Note: [1] These authors contributed equally to this work.

Abstract: 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.

Keywords: Aβ PET, amnestic mild cognitive impairment, Aβ positivity, machine learning, magnetic resonance imaging features, neuropsychological tests, prediction model

DOI: 10.3233/JAD-201092

Journal: Journal of Alzheimer's Disease, vol. 80, no. 1, pp. 143-157, 2021