Machine Learning Classification of Alzheimer’s Disease Stages Using Cerebrospinal Fluid Biomarkers Alone - Slides

Aug 4, 2026

Upload by Violet

Early diagnosis of Alzheimer’s disease remains difficult because common approaches fail to identify patients during preclinical stages that may persist for up to a decade before symptoms appear. This study applies machine learning to classify Alzheimer’s stages using cerebrospinal fluid biomarker levels only, leveraging amyloid beta 1-42, T-tau, and P-tau. Patient cohorts from the National Alzheimer’s Coordinating Centre database were subdivided with mini-mental state scores and clinical dementia ratings. Models including KNN, boosted/bagged trees, SVM, logistic regression, and Naïve Bayes were evaluated via statistical and correlation analyses to support stage separation and monitoring decisions.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages32

Reading Time1 h 21 min

AccessFree to download

Abstract

Introduction

Background on biomarkers and staging

Biomarker-based classification approach

Machine Learning Classification of Alzheimer’s Disease Stages Using Cerebrospinal Fluid Biomarkers Alone Vivek Kumar Tiwari1, Premananda Indic1, Shawana Tabassum1\# 1Department of Electrical Engineering, University of Texas, Tyler; Texas, U.S.A. \#Corresponding Author: Shawana Tabassum: [stabassum@uttyler.edu](stabassum@uttyler.edu) Abstract: Early diagnosis of Alzheimer’s disease is a challenge because the existing methodologies do not identify the patients in their preclinical stage, which can last up to a decade prior to the onset of clinical symptoms. Several research studies demonstrate the potential of cerebrospinal fluid biomarkers, amyloid beta 1-42, T-tau, and P-tau, in early diagnosis of Alzheimer’s disease stages. In this work, we used machine learning models to classify different stages of Alzheimer’s disease based on the cerebrospinal fluid biomarker levels alone. An electronic health record of patients from the National Alzheimer's Coordinating Centre database was analyzed and the patients were subdivided based on mini-mental state scores and clinical dementia ratings. Statistical and correlation analyses were performed to identify significant differences between the Alzheimer’s stages. Afterward, machine learning classifiers including K-Nearest Neighbors, Ensemble Boosted Tree, Ensemble Bagged Tree, Support Vector Machine, Logistic Regression, and Naïve Bayes classifiers were employed to classify the Alzheimer’s disease stages. The results demonstrate that Ensemble Boosted Tree (84.4%) and Logistic Regression (73.4%) provide the highest accuracy for binary classification, while Ensemble Bagged Tree (75.4%) demonstrates better accuracy for multiclassification. The findings from this research are expected to help clinicians in making an informed decision regarding the early diagnosis of Alzheimer’s from the cerebrospinal fluid biomarkers alone, monitoring of the disease progression, and implementation of appropriate intervention measures. Introduction Alzheimer’s disease (AD) is one of the most prevalent neurodegenerative diseases characterized by progressive cognitive decline. This cognitive deterioration also poses a significant burden on the caregivers. Despite the advances in disease interventions, the gap in knowledge for diagnosis and effective management of AD is threefold. First, biochemicals drive most physiological signals, yet the search for molecular biomarkers for precise and early detection of the pathology of AD remains a challenge. Biomarkers are biological molecules (for instance proteins) found in body fluids and indicate the stage of disease as well as the response to treatment. The well-established biomarkers of dementia include amyloid-beta peptide 1-42 (Aβ1-42), total tau (T-tau), and tau phosphorylated at threonine (P-tau) [1]. However, much is yet to be understood regarding the role of these molecular biomarkers in early diagnosis and decision-making for AD. Second, AD is often not diagnosed at an early stage because of the lack of a screening tool that can monitor/track the pathophysiological processes at the bedside. It is evident from positron emission tomography (PET) imaging studies that the underlying progressive pathology precedes the symptomatic onset of AD by one or two decades [2] . The underlying disease process likely begins years before AD manifests clinically [3] . As a result, significant irreversible neurological damage occurs by the time dementia is diagnosed, thereby reinforcing the need for point-of-care diagnosis for monitoring the pathogenesis ofAD and identifying the patients in predementia stages. Third, there is a lack of a classification framework for analyzing the time-series biomarker levels and making an informed decision about timely diagnosis and clinical interventions. The mild impairment of episodic memory is typically one of the first indications of patients with early-stage AD. These people may meet the requirements for mild cognitive impa

FAQs

Why is early diagnosis of Alzheimer’s disease challenging in the preclinical stage?

Existing methodologies often cannot identify patients before clinical symptoms, even though preclinical changes may last years to a decade. Early detection is critical to prevent irreversible neurological damage by the time dementia is diagnosed.

Which cerebrospinal fluid biomarkers and clinical scores are used in this work?

The study uses cerebrospinal fluid biomarkers amyloid beta 1-42, T-tau, and P-tau. Patients are subdivided using mini-mental state examination (MMSE) scores and clinical dementia rating (CDR) to define stages such as NC, MCI, and SD.

Which machine learning models performed best for binary and multiclass classification?

Ensemble Boosted Tree achieved the highest accuracy (84.4%) for binary classification, while Logistic Regression reached 73.4%. For multiclass classification, Ensemble Bagged Tree showed better accuracy at 75.4%.

Aug 4, 2026

Upload by Violet