Scholarlink

<p>Control of malaria depends on how soon and carefully the illness is treated. Many deep learning models, especially convolutional neural networks (CNNs), are now being used to improve image-based visual diagnosis. While these models can recognise malaria cases and affected areas, they rarely pinpoint specific places where malaria occurs. To address these difficulties, we introduced a deep learning method using CNN features and attention to achieve better disease detection and localisation accuracy. Specifically, we incorporate the Convolutional Block Attention Module (CBAM), which sequentially applies channel and spatial attention, allowing the model to focus on the most relevant regions in malaria images. We evaluated our model under identical conditions against established architectures, including XceptionNet, LeNet, MobileNet, NasNet, VGG16, and VGG19. We used precision, accuracy, F1-score, and recall to measure the performance. Our framework was better than the baseline models and scored 96% accuracy, 96% precision, 97% recall and 96% F1-score. Adding CBAM made it easier for the model to identify areas with malaria, raised diagnostic performance, and kept evaluation scores high. Lastly, we integrated Explainable AI (XAI) for our proposed model. The solution enables clinical staff to decide correctly on malaria management by offering an accurate and flexible malaria diagnosis approach. Our process provides both precise diagnosis and stronger identification of malaria in images, which makes it more useful for medical use.</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

Hybrid CNN Model with CBAM for Malaria Disease Classification: Enhanced Feature Extraction

  • 期刊论文

  • 发表日期：

    2025年12月8日

查看全文
  • Sayed Sayem,
  • Sayed Sumsul Islam Sanny,
  • Hadiur Rahman Nabil

摘要

Control of malaria depends on how soon and carefully the illness is treated. Many deep learning models, especially convolutional neural networks (CNNs), are now being used to improve image-based visual diagnosis. While these models can recognise malaria cases and affected areas, they rarely pinpoint specific places where malaria occurs. To address these difficulties, we introduced a deep learning method using CNN features and attention to achieve better disease detection and localisation accuracy. Specifically, we incorporate the Convolutional Block Attention Module (CBAM), which sequentially applies channel and spatial attention, allowing the model to focus on the most relevant regions in malaria images. We evaluated our model under identical conditions against established architectures, including XceptionNet, LeNet, MobileNet, NasNet, VGG16, and VGG19. We used precision, accuracy, F1-score, and recall to measure the performance. Our framework was better than the baseline models and scored 96% accuracy, 96% precision, 97% recall and 96% F1-score. Adding CBAM made it easier for the model to identify areas with malaria, raised diagnostic performance, and kept evaluation scores high. Lastly, we integrated Explainable AI (XAI) for our proposed model. The solution enables clinical staff to decide correctly on malaria management by offering an accurate and flexible malaria diagnosis approach. Our process provides both precise diagnosis and stronger identification of malaria in images, which makes it more useful for medical use.