Scholarlink

<p>Lung cancer is one of the primary diseases with a higher rate of death globally and it is caused because of the instinctive growth of abnormal cells in the lungs. The timely recognition of cancer in the lungs can save the lives of patients, who are affected by the disease of lung cancer. Numerous machines-based automated approaches, and image processing methods have been developed to detect lung cancer, but those methods face various challenges, such as computational complexities, overfitting, lower accuracy, and higher execution time in detecting lung cancer. Hence, this research proposed a Vulpes Hunt Optimization enabled Value Advantage Dense Implicit Triplet Attention-based 3D-Convolutional Gated Network (VHO-VADITA-3DCGN) model to detect lung cancer precisely. The integration of the Dense Implicit Triplet Attention (DITA) mechanism in the 3D-convolutional Gated Network (3DCGN) model helps to increase its accuracy in the detection process by improving the circulation of data over all the layers in the 3DCGN model. Additionally, the incorporation of Value Advantage (VA) learning in the model aids in accelerating the learning mechanism of the DITA-3DCGN model. Moreover, the Vulpes Hunt Optimization (VHO) algorithm is used to tune the learning parameters of the DITA-3DCGN by eliminating the issues of lower convergence rate and helps to enhance the overall performance of the model in detecting lung cancer. The VHO-VADITA-3DCGN model conquers an accuracy of 97.38%, <i>F</i>1-score of 96.79%, precision of 94.82%, and recall of 98.84% for K-Fold 10, respectively.</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

VADITA-3DCGN: Hybrid Attention-Based Optimized 3D-Convolutional Gated Network for Lung Cancer Detection

  • 期刊论文

  • 发表日期：

    2025年8月4日

查看全文
  • Bhavesh B Digey,
  • Shirish Shankar Kulkarni

摘要

Lung cancer is one of the primary diseases with a higher rate of death globally and it is caused because of the instinctive growth of abnormal cells in the lungs. The timely recognition of cancer in the lungs can save the lives of patients, who are affected by the disease of lung cancer. Numerous machines-based automated approaches, and image processing methods have been developed to detect lung cancer, but those methods face various challenges, such as computational complexities, overfitting, lower accuracy, and higher execution time in detecting lung cancer. Hence, this research proposed a Vulpes Hunt Optimization enabled Value Advantage Dense Implicit Triplet Attention-based 3D-Convolutional Gated Network (VHO-VADITA-3DCGN) model to detect lung cancer precisely. The integration of the Dense Implicit Triplet Attention (DITA) mechanism in the 3D-convolutional Gated Network (3DCGN) model helps to increase its accuracy in the detection process by improving the circulation of data over all the layers in the 3DCGN model. Additionally, the incorporation of Value Advantage (VA) learning in the model aids in accelerating the learning mechanism of the DITA-3DCGN model. Moreover, the Vulpes Hunt Optimization (VHO) algorithm is used to tune the learning parameters of the DITA-3DCGN by eliminating the issues of lower convergence rate and helps to enhance the overall performance of the model in detecting lung cancer. The VHO-VADITA-3DCGN model conquers an accuracy of 97.38%, F1-score of 96.79%, precision of 94.82%, and recall of 98.84% for K-Fold 10, respectively.