Using Machine Learning tools to Calculate Multi Slice Multi Echo (MSME) Score for Alzheimer's Diagnosis - Research Report

Aug 3, 2026

Upload by Miles

Alzheimer's disease (AD) is a major public health challenge, and the hippocampus offers a practical biomarker for MRI-based diagnosis in machine learning workflows. Prior approaches that rely on full MRI slices for AD classification have delivered limited accuracy. This study proposes a novel “select slices” method that focuses on specific hippocampal landmarks to remove irrelevant image information. Experiments use the ADNI dataset with ResNet50 and LeNet across sagittal, coronal, and axial views and normal/mild/severe categories.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages12

Reading Time30 min

AccessFree to download

Abstract

Introduction

Literature Review

Using Machine Learning tools to Calculate Multi Slice Multi Echo (MSME) Score for Alzheimer's Diagnosis SreedharYalamati Solutions Architect, Celer Systems Inc., Technology Services, CA, USA,0009-0009-4504-1467 1Received: 12 October 2023; Accepted: 26 December 2023; Published: 09 January 2024 ABSTRACT Alzheimer's disease (AD) poses a significant public health challenge. The hippocampus is one of the most affected brain regions and a readily accessible biomarker for diagnosis through MRI imaging in machine learning applications. However, utilizing entire MRI image slices in machine learning for AD classification has shown reduced accuracy. This study introduces the novel 'select slices' method, which involves identifying and focusing on specific landmarks within the hippocampus region in MRI images. This approach aims to improve classification accuracy by eliminating irrelevant information from the analysis. Our research aims to identify which views of MRI images produce higher accuracy for AD classification. We used multiclass classification using the publicly available Alzheimer's Disease Neuroimaging Initiative (ADNI) dataset, utilising Resnet50 and LeNet models to evaluate the value of three specific views (sagittal, coronal, and axial) and categories (normal, mild, and severe AD) . The dataset comprised 4,500 MRI slices across these three views and categories. Our findings unequivocally demonstrate that the 'select slices' approach surpasses the use of entire slices in MRI images for AD classification. Specifically, our method elevates machine learning accuracy, with the coronal view showcasing exceptional performance. This methodology is a game-changer in improving the accuracy ofmachine learning models for AD classification, with results closely mirroring those of medical experts, the gold standard for AD diagnosis. Moreover, we observed that LeNet models hold great promise as effective tools for AD classification. INTRODUCTION Alzheimer's disease (AD) stands as a pressing concern in public health, with approximately 44 million cases worldwide, a figure projected to soar to 131.5 million by 2050. This neurodegenerative ailment primarily afflicts individuals aged 65 and above, impairing memory and cognitive function. While AD lacks a definitive cure, certain medications and treatments can alleviate symptoms temporarily or slow its progression. Early diagnosis significantly eases the management of the condition. Early detection hinges on monitoring and investigating brain deterioration before AD advances. The hippocampus, a pivotal brain region, is a notable biomarker for AD due to its susceptibility to degeneration, particularly impacting memory. Magnetic resonance imaging (MRI) facilitates detecting hippocampal volume changes, aiding AD diagnosis. With its advanced imaging capabilities, MRI plays a pivotal role in analysing structural brain alterations. Advancements in machine learning have propelled the development of models capable of diagnosing AD based on MRI scans. These models excel in deciphering intricate patterns from MRI images, offering swift and accurate diagnoses. Leveraging machine learning expedites diagnosis compared to manual methods, ensuring consistency across vast datasets. Multiclass classification, distinguishing AD into three categories (AD, Mild Cognitive Impairment (MCI), and Normal Control (NC)), proves beneficial in capturing subtle disease progression nuances. 1 How to cite the article: Yalamati S., January 2024; Using Machine Learning tools to Calculate Multi Slice Multi Echo (MSME) Score for Alzheimer's Diagnosis; International Journal of Innovations in Scientific Engineering, Jan-Jun 2024, Vol 19, 1-12 Various strategies have been proposed to enhance AD classification accuracy, including improving MRI image quality, employing segmentation, and utilizing classification techniques such asAdaBoost. Selecting specific MRI slices rather than analysing entire datasets ha

FAQs

Why does using entire MRI slices reduce accuracy for AD classification?

Full-slice analysis can include irrelevant information, which weakens the model’s ability to learn discriminative patterns tied to hippocampal degeneration.

What is the “select slices” method proposed in this study?

The method identifies and focuses on specific landmarks within the hippocampus region, analyzing selected information rather than entire MRI slices.

Which MRI view and model approach showed the strongest results in the study?

The coronal view demonstrated exceptional performance, and the LeNet model also showed strong promise for AD classification.

Aug 3, 2026

Upload by Miles