Accurate and robust prediction of Amyloid-β brain deposition from plasma biomarkers and clinical information using machine learning

BackgroundAlzheimer's disease (AD) greatly affects the daily functioning and life quality of patients and is prevalent in the elderly population. Amyloid-β (Aβ) accumulation in the brain is the main hallmark of AD pathophysiology. Positron Emission Tomography (PET) imaging is the most accurate...

Full description

Saved in:
Bibliographic Details
Main Authors: Jiayuan Xu, Andrew J. Doig, Sofia Michopoulou, Petroula Proitsi, Fumie Costen, The Alzheimer's disease neuroimaging initiative
Format: Article
Language:English
Published: Frontiers Media S.A. 2025-08-01
Series:Frontiers in Aging Neuroscience
Subjects:
Online Access:https://www.frontiersin.org/articles/10.3389/fnagi.2025.1559459/full
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849340537269649408
author Jiayuan Xu
Andrew J. Doig
Sofia Michopoulou
Sofia Michopoulou
Petroula Proitsi
Petroula Proitsi
Fumie Costen
The Alzheimer's disease neuroimaging initiative
author_facet Jiayuan Xu
Andrew J. Doig
Sofia Michopoulou
Sofia Michopoulou
Petroula Proitsi
Petroula Proitsi
Fumie Costen
The Alzheimer's disease neuroimaging initiative
author_sort Jiayuan Xu
collection DOAJ
description BackgroundAlzheimer's disease (AD) greatly affects the daily functioning and life quality of patients and is prevalent in the elderly population. Amyloid-β (Aβ) accumulation in the brain is the main hallmark of AD pathophysiology. Positron Emission Tomography (PET) imaging is the most accurate method to identify Aβ deposits in the brain, but it is expensive and not widely available. The development of a low-cost method to detect Aβ deposition in the brain, as an alternative to PET, would therefore be of great value. This study aims to develop and validate machine learning algorithms for accurately predicting brain Aβ positivity using plasma biomarkers, genetic information, and clinical data as a cost-effective alternative to PET imaging.MethodsWe analyzed 1,043 patients from the Alzheimer's Disease Neuroimaging Initiative (ADNI) dataset and validated our models on 127 patients from the Center for Neurodegeneration and Translational Neuroscience (CNTN) dataset. Brain Aβ status was determined using plasma biomarkers [Aβ42, Aβ40, Phosphorylated tau (pTau) 181, Neurofilament light chain (NfL)], Apolipoprotein E (APOE) genotype, and clinical information [Mini-Mental State Examination (MMSE), Montreal Cognitive Assessment (MoCA), age, education year, and gender]. Decision tree, random forest, support vector machine, and multilayer perceptron machine learning methods were used to combine all this information. We introduced a feature selection method to balance the performance and the number of features. We conducted a feature matching technique to enable our model to be tested on the external dataset without retraining.ResultsOur system achieved a value of 0.95 for the Area Under the ROC curve (AUC) using the ADNI dataset (n = 340) and the full set of 11 features. Our architecture was also tested on an external dataset (CNTN, n = 127) and achieved an AUC of 0.90. When using only five features (pTau 181, Aβ42/40, Aβ42, APOE ɛ4 count, and MMSE) on 341 ADNI patients, we achieved an AUC of 0.87.ConclusionThe random forest, support vector machine and multilayer perceptron methods can accurately predict brain Aβ status using plasma biomarkers, genotype, and clinical information. The method generalizes well to an independent dataset and can be reduced to using only five features without losing much accuracy, thus providing an inexpensive alternative to PET imaging.
format Article
id doaj-art-5ed08c565c124779b7559cd6f704206c
institution Kabale University
issn 1663-4365
language English
publishDate 2025-08-01
publisher Frontiers Media S.A.
record_format Article
series Frontiers in Aging Neuroscience
spelling doaj-art-5ed08c565c124779b7559cd6f704206c2025-08-20T03:43:54ZengFrontiers Media S.A.Frontiers in Aging Neuroscience1663-43652025-08-011710.3389/fnagi.2025.15594591559459Accurate and robust prediction of Amyloid-β brain deposition from plasma biomarkers and clinical information using machine learningJiayuan Xu0Andrew J. Doig1Sofia Michopoulou2Sofia Michopoulou3Petroula Proitsi4Petroula Proitsi5Fumie Costen6The Alzheimer's disease neuroimaging initiativeDepartment of Electrical and Electronic Engineering, University of Manchester, Manchester, United KingdomDivision of Neuroscience, Stopford Building, School of Biological Sciences, Faculty of Biology, Medicine and Health, University of Manchester, Manchester, United KingdomMedical Physics University Hospital Southampton NHS Foundation Trust, Southampton, United KingdomClinical Experimental Sciences, University of Southampton, Southampton General Hospital, Southampton, United KingdomCentre for Preventive Neurology, Wolfson Institute of Population Health, Queen Mary's University of London, London, United KingdomDepartment of Basic and Clinical Neuroscience, Institute of Psychiatry, Psychology, and Neuroscience, King's College London, London, United KingdomDepartment of Electrical and Electronic Engineering, University of Manchester, Manchester, United KingdomBackgroundAlzheimer's disease (AD) greatly affects the daily functioning and life quality of patients and is prevalent in the elderly population. Amyloid-β (Aβ) accumulation in the brain is the main hallmark of AD pathophysiology. Positron Emission Tomography (PET) imaging is the most accurate method to identify Aβ deposits in the brain, but it is expensive and not widely available. The development of a low-cost method to detect Aβ deposition in the brain, as an alternative to PET, would therefore be of great value. This study aims to develop and validate machine learning algorithms for accurately predicting brain Aβ positivity using plasma biomarkers, genetic information, and clinical data as a cost-effective alternative to PET imaging.MethodsWe analyzed 1,043 patients from the Alzheimer's Disease Neuroimaging Initiative (ADNI) dataset and validated our models on 127 patients from the Center for Neurodegeneration and Translational Neuroscience (CNTN) dataset. Brain Aβ status was determined using plasma biomarkers [Aβ42, Aβ40, Phosphorylated tau (pTau) 181, Neurofilament light chain (NfL)], Apolipoprotein E (APOE) genotype, and clinical information [Mini-Mental State Examination (MMSE), Montreal Cognitive Assessment (MoCA), age, education year, and gender]. Decision tree, random forest, support vector machine, and multilayer perceptron machine learning methods were used to combine all this information. We introduced a feature selection method to balance the performance and the number of features. We conducted a feature matching technique to enable our model to be tested on the external dataset without retraining.ResultsOur system achieved a value of 0.95 for the Area Under the ROC curve (AUC) using the ADNI dataset (n = 340) and the full set of 11 features. Our architecture was also tested on an external dataset (CNTN, n = 127) and achieved an AUC of 0.90. When using only five features (pTau 181, Aβ42/40, Aβ42, APOE ɛ4 count, and MMSE) on 341 ADNI patients, we achieved an AUC of 0.87.ConclusionThe random forest, support vector machine and multilayer perceptron methods can accurately predict brain Aβ status using plasma biomarkers, genotype, and clinical information. The method generalizes well to an independent dataset and can be reduced to using only five features without losing much accuracy, thus providing an inexpensive alternative to PET imaging.https://www.frontiersin.org/articles/10.3389/fnagi.2025.1559459/fullAlzheimer's diseaseAβ PETplasma biomarkersmachine learning classification algorithmfeature selectionfeature matching
spellingShingle Jiayuan Xu
Andrew J. Doig
Sofia Michopoulou
Sofia Michopoulou
Petroula Proitsi
Petroula Proitsi
Fumie Costen
The Alzheimer's disease neuroimaging initiative
Accurate and robust prediction of Amyloid-β brain deposition from plasma biomarkers and clinical information using machine learning
Frontiers in Aging Neuroscience
Alzheimer's disease
Aβ PET
plasma biomarkers
machine learning classification algorithm
feature selection
feature matching
title Accurate and robust prediction of Amyloid-β brain deposition from plasma biomarkers and clinical information using machine learning
title_full Accurate and robust prediction of Amyloid-β brain deposition from plasma biomarkers and clinical information using machine learning
title_fullStr Accurate and robust prediction of Amyloid-β brain deposition from plasma biomarkers and clinical information using machine learning
title_full_unstemmed Accurate and robust prediction of Amyloid-β brain deposition from plasma biomarkers and clinical information using machine learning
title_short Accurate and robust prediction of Amyloid-β brain deposition from plasma biomarkers and clinical information using machine learning
title_sort accurate and robust prediction of amyloid β brain deposition from plasma biomarkers and clinical information using machine learning
topic Alzheimer's disease
Aβ PET
plasma biomarkers
machine learning classification algorithm
feature selection
feature matching
url https://www.frontiersin.org/articles/10.3389/fnagi.2025.1559459/full
work_keys_str_mv AT jiayuanxu accurateandrobustpredictionofamyloidbbraindepositionfromplasmabiomarkersandclinicalinformationusingmachinelearning
AT andrewjdoig accurateandrobustpredictionofamyloidbbraindepositionfromplasmabiomarkersandclinicalinformationusingmachinelearning
AT sofiamichopoulou accurateandrobustpredictionofamyloidbbraindepositionfromplasmabiomarkersandclinicalinformationusingmachinelearning
AT sofiamichopoulou accurateandrobustpredictionofamyloidbbraindepositionfromplasmabiomarkersandclinicalinformationusingmachinelearning
AT petroulaproitsi accurateandrobustpredictionofamyloidbbraindepositionfromplasmabiomarkersandclinicalinformationusingmachinelearning
AT petroulaproitsi accurateandrobustpredictionofamyloidbbraindepositionfromplasmabiomarkersandclinicalinformationusingmachinelearning
AT fumiecosten accurateandrobustpredictionofamyloidbbraindepositionfromplasmabiomarkersandclinicalinformationusingmachinelearning
AT thealzheimersdiseaseneuroimaginginitiative accurateandrobustpredictionofamyloidbbraindepositionfromplasmabiomarkersandclinicalinformationusingmachinelearning