Early Detection of Alzheimer’s Disease From Cortical and Hippocampal Local Field Potentials Using an Ensembled Machine Learning Model - Research paper

Aug 5, 2026

Upload by Patrick

Early diagnosis of Alzheimer’s disease remains difficult because spontaneous neuronal signals are complex to decode, even when multimodal data is used. This study proposes an explainable ensembled machine learning model (EXML) to detect subtle early-stage markers in cortical and hippocampal local field potentials (LFPs) from healthy and two AD animal model groups. LFP features are derived from temporal, spatial, and spectral domains and fused with late fusion, reaching 99.4% accuracy. Results also remain robust under channel masking to simulate artifacts and support mechanistic insights into amyloid plaque deposition.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages10

Reading Time25 min

AccessFree to download

Introduction

Motivation for early diagnosis

Data-driven and machine learning approaches

Role of biomarkers and neural signals

Materials and Methods

Explainable ensembled ML model (EXML)

LFP feature generation across domains

Late fusion strategy and robustness testing

Results and Discussion

Detection performance and accuracy

Insights into early amyloid plaque deposition

Robustness under masked channels

IEEE TRANSACTIONS ON NEURAL SYSTEMS AND REHABILITATION ENGINEERING, VOL. 31, 2023 2839 Early Detection of Alzheimer’s Disease From Cortical and Hippocampal Local Field Potentials Using an Ensembled Machine Learning Model Marcos Fabietti , Member, IEEE, Mufti Mahmud , Senior Member, IEEE, Ahmad Lotfi , Senior Member, IEEE, Alessandro Leparulo, Roberto Fontana , Stefano Vassanelli, and Cristina Fasolato Abstract—Early diagnosis of Alzheimer’s disease (AD) is a very challenging problem and has been attempted through data-driven methods in recent years. However, considering the inherent complexity in decoding higher cognitive functions from spontaneous neuronal signals, these data-driven methods benefit from the incorporation of multimodal data. This work proposes an ensembled machine learning model with explainability (EXML) to detect subtle patterns in cortical and hippocampal local field potential signals (LFPs) that can be considered as a potential marker for AD in the early stage of the disease. The LFPs acquired from healthy and two types of AD animal models (n = 10 each) using linear multielectrode probes were endorsed by electrocardiogram and respiration signals for their veracity. Feature sets were generated from LFPs in temporal, spatial and spectral domains and were fed into selected machine-learning models for each domain. Using late fusion, the EXML model achieved an overall accuracy of 99.4%. This provided insights into the amyloid plaque deposition process as early as 3 months of the disease onset by identifying the subtle patterns in the network activities. Lastly, the individual and ensemble models were found to be robust when evaluated by randomly masking channels to mimic the presence of artefacts. Manuscript received 3 December 2022; revised 29 March 2023; accepted 26 April 2023 . Date of publication 22 June 2023; date of current version 6 July 2023 . Marcos Fabietti was supported by the Nottingham Trent University Vice Chancellor’s Doctoral Fellowship. Mufti Mahmud was supported by Nottingham Trent University’s Strategic Research Theme Springboard Fund. The data collection was supported by the the Italian Ministry of Education, Universities and Research (MIUR) through PRIN-20175C22WM grant to Cristina Fasolato.(Corresponding author: Mufti Mahmud.) This work involved human subjects or animals in its research. Approval of all ethical and experimental procedures and protocols was granted by the Animal Care Committee of the University of Padua and the Italian Ministry of Health under Application No. 522/2018-PR. Marcos Fabietti and Ahmad Lotfi are with the Department of Computer Science, Nottingham Trent University, NG11 8NS Nottingham, U.K. Mufti Mahmud is with the Department of Computer Science, Nottingham Trent University, NG11 8NS Nottingham, U.K., and also with the Medical Technologies Innovation Facility and the Computing and Informatics Research Centre of Nottingham Trent University, NG11 8NS Nottingham, U.K. (e-mail: [mufti.mahmud@ntu.ac.uk](mufti.mahmud@ntu.ac.uk) ; [muftimahmud@gmail.com](muftimahmud@gmail.com)). Alessandro Leparulo, Stefano Vassanelli, and Cristina Fasolato are with the Department of Biomedical Sciences, University of Padova, 35127 Padova, Italy. Roberto Fontana is with the Department of Physiology and Pharmacology, Sapienza Università di Roma, 00185 Rome, Italy. Digital Object Identifier 10.1109/TNSRE.2023.3288835 Index Terms—Deep learning, dementia, neuronal signals, neuronal network, multimodal. I. INTRODUCTION ALZHEIMER’S disease (AD) is a neuropathology, which affects 46.8 million people worldwide [1] . While the disease is not fully understood, it is known that the lesions caused by amyloid plaques and tangles disrupt the connectivity of neurons, leading to their death and the atrophy of the brain. The main consequence is dementia, which impacts the person’s ability to think, behave, work and function independently. The total healthcare cost for the tre

FAQs

What is the main goal of this study on Alzheimer’s detection?

To detect subtle early-stage patterns associated with Alzheimer’s disease using cortical and hippocampal local field potentials and an explainable ensembled machine learning model.

How does the EXML model use neural signals for prediction?

It generates feature sets from LFP signals in temporal, spatial, and spectral domains, then feeds them into selected machine-learning models and combines outcomes using late fusion.

What performance and robustness does the model achieve?

The late-fusion EXML model achieves overall accuracy of 99.4% and remains robust when evaluated by randomly masking channels to mimic artifacts.

Aug 5, 2026

Upload by Patrick