ViTAD:利用改进的视觉转换器从脑磁共振成像扫描中对阿尔茨海默病进行多阶段分类。

IF 2.7 4区 医学 Q3 NEUROSCIENCES Brain Research Pub Date : 2024-11-12 DOI:10.1016/j.brainres.2024.149302
Md. Ashif Mahmud Joy , Shamima Nasrin , Ayesha Siddiqua , Dewan Md. Farid
{"title":"ViTAD:利用改进的视觉转换器从脑磁共振成像扫描中对阿尔茨海默病进行多阶段分类。","authors":"Md. Ashif Mahmud Joy ,&nbsp;Shamima Nasrin ,&nbsp;Ayesha Siddiqua ,&nbsp;Dewan Md. Farid","doi":"10.1016/j.brainres.2024.149302","DOIUrl":null,"url":null,"abstract":"<div><div>Alzheimer’s disease (AD) is a progressive neurological disorder that significantly impairs cognitive functions, particularly memory and thinking skills. The presence of AD in millions of individuals worldwide constitutes a substantial global health challenge. Timely and accurate diagnosis of AD is critical for effective management and improved patient outcomes. This study introduces ViTAD, an innovative method for classifying five stages of AD from brain MRI images, leveraging a Vision Transformer (ViT) model. The proposed model modifies Google’s ViT architecture, incorporating fine-tuned hyperparameters and additional layers to enhance its performance for AD stage detection. The dataset comprises 1,296 brain MRI images from the ADNI dataset, covering five stages of AD: Cognitively Normal (CN), Early Mild Cognitive Impairment (EMCI), Late Mild Cognitive Impairment (LMCI), Mild Cognitive Impairment (MCI), and Alzheimer’s Disease (AD). Our preprocessing pipeline includes grayscale to RGB conversion, image cropping, and the application of a Laplacian sharpening filter to enhance image clarity. Data augmentation was performed using horizontal/vertical flips, zoom, and rotation to ensure model robustness. We allocated 85% of the dataset for training and 15% for testing. Upon training the model for 20 epochs with a learning rate of 0.0001, ViTAD achieved a remarkable 99.98% accuracy, with 100% precision and an F1-score of 1.00. ViTAD’s superior performance in the multi-class classification task outperforms several conventional CNN-based models such as DenseNet and EfficientNet, which struggled with the 5-class AD detection task. Additionally, ViTAD demonstrated high efficiency, achieving optimal accuracy within only 8 epochs, far surpassing traditional CNN models in speed and accuracy. These findings highlight the significant potential of ViTAD as an automated, accurate, and efficient tool for early diagnosis of AD, offering valuable support in clinical settings.</div></div>","PeriodicalId":9083,"journal":{"name":"Brain Research","volume":"1847 ","pages":"Article 149302"},"PeriodicalIF":2.7000,"publicationDate":"2024-11-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"ViTAD: Leveraging modified vision transformer for Alzheimer’s disease multi-stage classification from brain MRI scans\",\"authors\":\"Md. Ashif Mahmud Joy ,&nbsp;Shamima Nasrin ,&nbsp;Ayesha Siddiqua ,&nbsp;Dewan Md. Farid\",\"doi\":\"10.1016/j.brainres.2024.149302\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"<div><div>Alzheimer’s disease (AD) is a progressive neurological disorder that significantly impairs cognitive functions, particularly memory and thinking skills. The presence of AD in millions of individuals worldwide constitutes a substantial global health challenge. Timely and accurate diagnosis of AD is critical for effective management and improved patient outcomes. This study introduces ViTAD, an innovative method for classifying five stages of AD from brain MRI images, leveraging a Vision Transformer (ViT) model. The proposed model modifies Google’s ViT architecture, incorporating fine-tuned hyperparameters and additional layers to enhance its performance for AD stage detection. The dataset comprises 1,296 brain MRI images from the ADNI dataset, covering five stages of AD: Cognitively Normal (CN), Early Mild Cognitive Impairment (EMCI), Late Mild Cognitive Impairment (LMCI), Mild Cognitive Impairment (MCI), and Alzheimer’s Disease (AD). Our preprocessing pipeline includes grayscale to RGB conversion, image cropping, and the application of a Laplacian sharpening filter to enhance image clarity. Data augmentation was performed using horizontal/vertical flips, zoom, and rotation to ensure model robustness. We allocated 85% of the dataset for training and 15% for testing. Upon training the model for 20 epochs with a learning rate of 0.0001, ViTAD achieved a remarkable 99.98% accuracy, with 100% precision and an F1-score of 1.00. ViTAD’s superior performance in the multi-class classification task outperforms several conventional CNN-based models such as DenseNet and EfficientNet, which struggled with the 5-class AD detection task. Additionally, ViTAD demonstrated high efficiency, achieving optimal accuracy within only 8 epochs, far surpassing traditional CNN models in speed and accuracy. These findings highlight the significant potential of ViTAD as an automated, accurate, and efficient tool for early diagnosis of AD, offering valuable support in clinical settings.</div></div>\",\"PeriodicalId\":9083,\"journal\":{\"name\":\"Brain Research\",\"volume\":\"1847 \",\"pages\":\"Article 149302\"},\"PeriodicalIF\":2.7000,\"publicationDate\":\"2024-11-12\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Brain Research\",\"FirstCategoryId\":\"3\",\"ListUrlMain\":\"https://www.sciencedirect.com/science/article/pii/S0006899324005560\",\"RegionNum\":4,\"RegionCategory\":\"医学\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q3\",\"JCRName\":\"NEUROSCIENCES\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Brain Research","FirstCategoryId":"3","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S0006899324005560","RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q3","JCRName":"NEUROSCIENCES","Score":null,"Total":0}
引用次数: 0

摘要

阿尔茨海默病(AD)是一种渐进性神经系统疾病,严重损害认知功能,尤其是记忆和思维能力。全世界有数百万人患有阿尔茨海默病,这对全球健康构成了巨大挑战。及时准确地诊断出老年痴呆症对于有效管理和改善患者预后至关重要。本研究介绍了 ViTAD,这是一种利用视觉转换器(ViT)模型从脑磁共振成像图像中对注意力缺失症的五个阶段进行分类的创新方法。所提出的模型修改了谷歌的 ViT 架构,纳入了经过微调的超参数和附加层,以提高其在 AD 阶段检测方面的性能。数据集包括来自 ADNI 数据集的 1,296 张脑磁共振成像图像,涵盖 AD 的五个阶段:认知正常(CN)、早期轻度认知障碍(EMCI)、晚期轻度认知障碍(LMCI)、轻度认知障碍(MCI)和阿尔茨海默病(AD)。我们的预处理流程包括灰度到 RGB 转换、图像裁剪和应用拉普拉斯锐化滤波器来提高图像清晰度。为了确保模型的稳健性,我们使用水平/垂直翻转、缩放和旋转等方法对数据进行了增强。我们将数据集的 85% 用于训练,15% 用于测试。在以 0.0001 的学习率对模型进行了 20 次历时训练后,ViTAD 的准确率达到了 99.98%,精确度为 100%,F1 分数为 1.00。ViTAD 在多类分类任务中的优异表现优于 DenseNet 和 EfficientNet 等几种基于 CNN 的传统模型,而这些模型在 5 类 AD 检测任务中表现不佳。此外,ViTAD 还表现出很高的效率,只用了 8 个历时就达到了最佳准确率,在速度和准确率上都远远超过了传统的 CNN 模型。这些发现凸显了 ViTAD 作为一种自动、准确、高效的工具在早期诊断注意力缺失症方面的巨大潜力,为临床提供了宝贵的支持。
本文章由计算机程序翻译,如有差异,请以英文原文为准。

摘要图片

查看原文
分享 分享
微信好友 朋友圈 QQ好友 复制链接
本刊更多论文
ViTAD: Leveraging modified vision transformer for Alzheimer’s disease multi-stage classification from brain MRI scans
Alzheimer’s disease (AD) is a progressive neurological disorder that significantly impairs cognitive functions, particularly memory and thinking skills. The presence of AD in millions of individuals worldwide constitutes a substantial global health challenge. Timely and accurate diagnosis of AD is critical for effective management and improved patient outcomes. This study introduces ViTAD, an innovative method for classifying five stages of AD from brain MRI images, leveraging a Vision Transformer (ViT) model. The proposed model modifies Google’s ViT architecture, incorporating fine-tuned hyperparameters and additional layers to enhance its performance for AD stage detection. The dataset comprises 1,296 brain MRI images from the ADNI dataset, covering five stages of AD: Cognitively Normal (CN), Early Mild Cognitive Impairment (EMCI), Late Mild Cognitive Impairment (LMCI), Mild Cognitive Impairment (MCI), and Alzheimer’s Disease (AD). Our preprocessing pipeline includes grayscale to RGB conversion, image cropping, and the application of a Laplacian sharpening filter to enhance image clarity. Data augmentation was performed using horizontal/vertical flips, zoom, and rotation to ensure model robustness. We allocated 85% of the dataset for training and 15% for testing. Upon training the model for 20 epochs with a learning rate of 0.0001, ViTAD achieved a remarkable 99.98% accuracy, with 100% precision and an F1-score of 1.00. ViTAD’s superior performance in the multi-class classification task outperforms several conventional CNN-based models such as DenseNet and EfficientNet, which struggled with the 5-class AD detection task. Additionally, ViTAD demonstrated high efficiency, achieving optimal accuracy within only 8 epochs, far surpassing traditional CNN models in speed and accuracy. These findings highlight the significant potential of ViTAD as an automated, accurate, and efficient tool for early diagnosis of AD, offering valuable support in clinical settings.
求助全文
通过发布文献求助,成功后即可免费获取论文全文。 去求助
来源期刊
Brain Research
Brain Research 医学-神经科学
CiteScore
5.90
自引率
3.40%
发文量
268
审稿时长
47 days
期刊介绍: An international multidisciplinary journal devoted to fundamental research in the brain sciences. Brain Research publishes papers reporting interdisciplinary investigations of nervous system structure and function that are of general interest to the international community of neuroscientists. As is evident from the journals name, its scope is broad, ranging from cellular and molecular studies through systems neuroscience, cognition and disease. Invited reviews are also published; suggestions for and inquiries about potential reviews are welcomed. With the appearance of the final issue of the 2011 subscription, Vol. 67/1-2 (24 June 2011), Brain Research Reviews has ceased publication as a distinct journal separate from Brain Research. Review articles accepted for Brain Research are now published in that journal.
期刊最新文献
Memory is improved and reflex maturation accelerated in the progeny of rat dams that consumed pequi (Caryocar Brasiliense). DLGAP3 suppresses malignant behaviors of glioma cells via inhibiting RGS12-mediated MAPK/ERK signaling. Editorial Board Gamma oscillation optimally predicts finger movements Upstream regulation of microRNA-9 through a complex cellular machinery during neurogenesis
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
现在去查看 取消
×
提示
确定
0
微信
客服QQ
Book学术公众号 扫码关注我们
反馈
×
意见反馈
请填写您的意见或建议
请填写您的手机或邮箱
已复制链接
已复制链接
快去分享给好友吧!
我知道了
×
扫码分享
扫码分享
Book学术官方微信
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术
文献互助 智能选刊 最新文献 互助须知 联系我们:info@booksci.cn
Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。
Copyright © 2023 Book学术 All rights reserved.
ghs 京公网安备 11010802042870号 京ICP备2023020795号-1