Early-Stage Alzheimer’s Disease Categorization Using PET Neuroimaging Modality and Convolutional Neural Networks in the 2D and 3D Domains
Alzheimer’s Disease (AD) is a health apprehension of significant proportions that is negatively impacting the ageing population globally. It is characterized by neuronal loss and the formation of structures such as neurofibrillary tangles and amyloid plaques in the early as well as later stages of t...
Main Authors: | , , , , , , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2022-06-01
|
Series: | Sensors |
Subjects: | |
Online Access: | https://www.mdpi.com/1424-8220/22/12/4609 |
_version_ | 1797482351872180224 |
---|---|
author | Ahsan Bin Tufail Nazish Anwar Mohamed Tahar Ben Othman Inam Ullah Rehan Ali Khan Yong-Kui Ma Deepak Adhikari Ateeq Ur Rehman Muhammad Shafiq Habib Hamam |
author_facet | Ahsan Bin Tufail Nazish Anwar Mohamed Tahar Ben Othman Inam Ullah Rehan Ali Khan Yong-Kui Ma Deepak Adhikari Ateeq Ur Rehman Muhammad Shafiq Habib Hamam |
author_sort | Ahsan Bin Tufail |
collection | DOAJ |
description | Alzheimer’s Disease (AD) is a health apprehension of significant proportions that is negatively impacting the ageing population globally. It is characterized by neuronal loss and the formation of structures such as neurofibrillary tangles and amyloid plaques in the early as well as later stages of the disease. Neuroimaging modalities are routinely used in clinical practice to capture brain alterations associated with AD. On the other hand, deep learning methods are routinely used to recognize patterns in underlying data distributions effectively. This work uses Convolutional Neural Network (CNN) architectures in both 2D and 3D domains to classify the initial stages of AD into AD, Mild Cognitive Impairment (MCI) and Normal Control (NC) classes using the positron emission tomography neuroimaging modality deploying data augmentation in a random zoomed in/out scheme. We used novel concepts such as the blurring before subsampling principle and distant domain transfer learning to build 2D CNN architectures. We performed three binaries, that is, AD/NC, AD/MCI, MCI/NC and one multiclass classification task AD/NC/MCI. The statistical comparison revealed that 3D-CNN architecture performed the best achieving an accuracy of 89.21% on AD/NC, 71.70% on AD/MCI, 62.25% on NC/MCI and 59.73% on AD/NC/MCI classification tasks using a five-fold cross-validation hyperparameter selection approach. Data augmentation helps in achieving superior performance on the multiclass classification task. The obtained results support the application of deep learning models towards early recognition of AD. |
first_indexed | 2024-03-09T22:31:04Z |
format | Article |
id | doaj.art-91a88ea00bf0417fad0132c6567ef4c4 |
institution | Directory Open Access Journal |
issn | 1424-8220 |
language | English |
last_indexed | 2024-03-09T22:31:04Z |
publishDate | 2022-06-01 |
publisher | MDPI AG |
record_format | Article |
series | Sensors |
spelling | doaj.art-91a88ea00bf0417fad0132c6567ef4c42023-11-23T18:56:04ZengMDPI AGSensors1424-82202022-06-012212460910.3390/s22124609Early-Stage Alzheimer’s Disease Categorization Using PET Neuroimaging Modality and Convolutional Neural Networks in the 2D and 3D DomainsAhsan Bin Tufail0Nazish Anwar1Mohamed Tahar Ben Othman2Inam Ullah3Rehan Ali Khan4Yong-Kui Ma5Deepak Adhikari6Ateeq Ur Rehman7Muhammad Shafiq8Habib Hamam9School of Electronics and Information Engineering, Harbin Institute of Technology, Harbin 150001, ChinaRegistered Medical Practitioner, Pakistan Medical Commission, Islamabad 44000, PakistanDepartment of Computer Science, College of Computer, Qassim University, Buraydah 51452, Saudi ArabiaCollege of Internet of Things (IoT) Engineering, Changzhou Campus, Hohai University (HHU), Changzhou 213022, ChinaDepartment of Electrical Engineering, University of Science and Technology Bannu, Bannu 28100, PakistanSchool of Electronics and Information Engineering, Harbin Institute of Technology, Harbin 150001, ChinaSchool of Information and Software Engineering, University of Electronic Science and Technology of China, Chengdu 611731, ChinaDepartment of Electrical Engineering, Government College University Lahore, Lahore 54000, PakistanDepartment of Information and Communication Engineering, Yeungnam University, Gyeongsan 38541, KoreaFaculty of Engineering, Université de Moncton, Moncton, NB E1A3E9, CanadaAlzheimer’s Disease (AD) is a health apprehension of significant proportions that is negatively impacting the ageing population globally. It is characterized by neuronal loss and the formation of structures such as neurofibrillary tangles and amyloid plaques in the early as well as later stages of the disease. Neuroimaging modalities are routinely used in clinical practice to capture brain alterations associated with AD. On the other hand, deep learning methods are routinely used to recognize patterns in underlying data distributions effectively. This work uses Convolutional Neural Network (CNN) architectures in both 2D and 3D domains to classify the initial stages of AD into AD, Mild Cognitive Impairment (MCI) and Normal Control (NC) classes using the positron emission tomography neuroimaging modality deploying data augmentation in a random zoomed in/out scheme. We used novel concepts such as the blurring before subsampling principle and distant domain transfer learning to build 2D CNN architectures. We performed three binaries, that is, AD/NC, AD/MCI, MCI/NC and one multiclass classification task AD/NC/MCI. The statistical comparison revealed that 3D-CNN architecture performed the best achieving an accuracy of 89.21% on AD/NC, 71.70% on AD/MCI, 62.25% on NC/MCI and 59.73% on AD/NC/MCI classification tasks using a five-fold cross-validation hyperparameter selection approach. Data augmentation helps in achieving superior performance on the multiclass classification task. The obtained results support the application of deep learning models towards early recognition of AD.https://www.mdpi.com/1424-8220/22/12/4609Alzheimer’s diseasebinary classificationmulticlass classificationstatistical evaluationpositron emission tomographydeep learning |
spellingShingle | Ahsan Bin Tufail Nazish Anwar Mohamed Tahar Ben Othman Inam Ullah Rehan Ali Khan Yong-Kui Ma Deepak Adhikari Ateeq Ur Rehman Muhammad Shafiq Habib Hamam Early-Stage Alzheimer’s Disease Categorization Using PET Neuroimaging Modality and Convolutional Neural Networks in the 2D and 3D Domains Sensors Alzheimer’s disease binary classification multiclass classification statistical evaluation positron emission tomography deep learning |
title | Early-Stage Alzheimer’s Disease Categorization Using PET Neuroimaging Modality and Convolutional Neural Networks in the 2D and 3D Domains |
title_full | Early-Stage Alzheimer’s Disease Categorization Using PET Neuroimaging Modality and Convolutional Neural Networks in the 2D and 3D Domains |
title_fullStr | Early-Stage Alzheimer’s Disease Categorization Using PET Neuroimaging Modality and Convolutional Neural Networks in the 2D and 3D Domains |
title_full_unstemmed | Early-Stage Alzheimer’s Disease Categorization Using PET Neuroimaging Modality and Convolutional Neural Networks in the 2D and 3D Domains |
title_short | Early-Stage Alzheimer’s Disease Categorization Using PET Neuroimaging Modality and Convolutional Neural Networks in the 2D and 3D Domains |
title_sort | early stage alzheimer s disease categorization using pet neuroimaging modality and convolutional neural networks in the 2d and 3d domains |
topic | Alzheimer’s disease binary classification multiclass classification statistical evaluation positron emission tomography deep learning |
url | https://www.mdpi.com/1424-8220/22/12/4609 |
work_keys_str_mv | AT ahsanbintufail earlystagealzheimersdiseasecategorizationusingpetneuroimagingmodalityandconvolutionalneuralnetworksinthe2dand3ddomains AT nazishanwar earlystagealzheimersdiseasecategorizationusingpetneuroimagingmodalityandconvolutionalneuralnetworksinthe2dand3ddomains AT mohamedtaharbenothman earlystagealzheimersdiseasecategorizationusingpetneuroimagingmodalityandconvolutionalneuralnetworksinthe2dand3ddomains AT inamullah earlystagealzheimersdiseasecategorizationusingpetneuroimagingmodalityandconvolutionalneuralnetworksinthe2dand3ddomains AT rehanalikhan earlystagealzheimersdiseasecategorizationusingpetneuroimagingmodalityandconvolutionalneuralnetworksinthe2dand3ddomains AT yongkuima earlystagealzheimersdiseasecategorizationusingpetneuroimagingmodalityandconvolutionalneuralnetworksinthe2dand3ddomains AT deepakadhikari earlystagealzheimersdiseasecategorizationusingpetneuroimagingmodalityandconvolutionalneuralnetworksinthe2dand3ddomains AT ateequrrehman earlystagealzheimersdiseasecategorizationusingpetneuroimagingmodalityandconvolutionalneuralnetworksinthe2dand3ddomains AT muhammadshafiq earlystagealzheimersdiseasecategorizationusingpetneuroimagingmodalityandconvolutionalneuralnetworksinthe2dand3ddomains AT habibhamam earlystagealzheimersdiseasecategorizationusingpetneuroimagingmodalityandconvolutionalneuralnetworksinthe2dand3ddomains |