Scholarlink

<p>Alzheimer’s disease (AD) is a progressive neurological disorder characterized by memory loss, confusion, and behavioral changes, and is the leading cause of dementia, impairing daily functioning. As the disease progresses gradually from mild cognitive issues to severe decline, early and accurate detection is vital for effective intervention. However, existing diagnostic approaches often suffer from limited accuracy, high false positive rates, and poor interpretability. To overcome these limitations, this study proposes a novel diagnostic framework, Fuzzy Residual-LeNet (FuzzyRes-LeNet), paired with the newly developed Harmonic Fossa Optimization Algorithm (HarFOA), specifically designed to enhance segmentation precision and classification performance. The proposed framework begins with preprocessing Magnetic Resonance Images (MRIs) using a Gaussian filter to enhance image quality. This is followed by image segmentation using O-SegUNet, a hybrid architecture combining UNet and O-SegNet, trained with a focal loss function to address class imbalance. The training of O-SegUNet is optimized using HarFOA, which integrates harmonic analysis with the traditional Fossa Optimization Algorithm (FOA). After segmentation, relevant features are extracted and used for classification utilizing FuzzyRes-LeNet, a novel model that integrates LeNet, ResNeXt, and fuzzy logic, also fine-tuned using HarFOA. In addition, the developed FuzzyRes-LeNet_HarFOA attains better False Positive Rate (FPR), accuracy, and True Positive Rate (TPR) of 5.412%, 94.475%, and 94.999%.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Segmentation and Fuzzy Res-LeNet-Based Classification for Alzheimer's Disease Using Harmonic Fossa Optimization

  • 期刊论文

  • 发表日期：

    2025年9月28日

查看全文
  • R. Soujanya,
  • Maganti Syamala,
  • Sai Hanuman Akundi

摘要

Alzheimer’s disease (AD) is a progressive neurological disorder characterized by memory loss, confusion, and behavioral changes, and is the leading cause of dementia, impairing daily functioning. As the disease progresses gradually from mild cognitive issues to severe decline, early and accurate detection is vital for effective intervention. However, existing diagnostic approaches often suffer from limited accuracy, high false positive rates, and poor interpretability. To overcome these limitations, this study proposes a novel diagnostic framework, Fuzzy Residual-LeNet (FuzzyRes-LeNet), paired with the newly developed Harmonic Fossa Optimization Algorithm (HarFOA), specifically designed to enhance segmentation precision and classification performance. The proposed framework begins with preprocessing Magnetic Resonance Images (MRIs) using a Gaussian filter to enhance image quality. This is followed by image segmentation using O-SegUNet, a hybrid architecture combining UNet and O-SegNet, trained with a focal loss function to address class imbalance. The training of O-SegUNet is optimized using HarFOA, which integrates harmonic analysis with the traditional Fossa Optimization Algorithm (FOA). After segmentation, relevant features are extracted and used for classification utilizing FuzzyRes-LeNet, a novel model that integrates LeNet, ResNeXt, and fuzzy logic, also fine-tuned using HarFOA. In addition, the developed FuzzyRes-LeNet_HarFOA attains better False Positive Rate (FPR), accuracy, and True Positive Rate (TPR) of 5.412%, 94.475%, and 94.999%.