Scholarlink

<p>The most precise method for identifying lung cancer (LC), one of the fatal illnesses discovered in recent decades, is via CT scan pictures obtained after a biopsy. However, the limitations of current deep learning (DL) techniques in medical imaging often lead to inaccuracies causing misdiagnosis and inappropriate treatment. In this regard, the current research proposes a Pyramid Quantum Convolutional Neural Network with Migrating Walrus Algorithm (PQCNNet-MWA) for improved LC detection from CT scan pictures retrieved from IQ-OTH/NCCD as well as LIDC-IDRI databases. The proposed method initiates by preprocessing using Sub-Aperture Keystone Transform-Matched Filtering (SAKTMF), where enhancement and normalization of raw input data take place; segmentation follows, yielding Q-value regularized Transformer (Q-VRT) that segments images into meaningful regions for accurate analysis of the data. Classification is done using PQCNNet, combining Pyramid Attention (PA) and Quantum Convolutional Neural Networks (QCNN) for better feature extraction as well as classification performance; this process is optimized using the Migrating Walrus Algorithm for enhanced accuracy. The proposed approach obtained 99.9% accuracy as well as 99.8% precision for the IQ-OTH/NCCD database and 99.8% accuracy and 99.7% precision for the LIDC-IDRI database. These results show that the model works better than existing methods, suggesting a major advancement in LC diagnosis.</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

Effective Diagnosis of Lung Cancer Using Pyramid Quantum Convolutional Neural Network with Migrating Walrus Algorithm on CT Scan Images

  • 期刊论文

  • 发表日期：

    2025年9月10日

查看全文
  • Seema Ravindra Baji,
  • Sahebrao Bhaurao Bagal,
  • Sachin Vasant Chaudhari,
  • Balasaheb Shrirangrao Agarkar

摘要

The most precise method for identifying lung cancer (LC), one of the fatal illnesses discovered in recent decades, is via CT scan pictures obtained after a biopsy. However, the limitations of current deep learning (DL) techniques in medical imaging often lead to inaccuracies causing misdiagnosis and inappropriate treatment. In this regard, the current research proposes a Pyramid Quantum Convolutional Neural Network with Migrating Walrus Algorithm (PQCNNet-MWA) for improved LC detection from CT scan pictures retrieved from IQ-OTH/NCCD as well as LIDC-IDRI databases. The proposed method initiates by preprocessing using Sub-Aperture Keystone Transform-Matched Filtering (SAKTMF), where enhancement and normalization of raw input data take place; segmentation follows, yielding Q-value regularized Transformer (Q-VRT) that segments images into meaningful regions for accurate analysis of the data. Classification is done using PQCNNet, combining Pyramid Attention (PA) and Quantum Convolutional Neural Networks (QCNN) for better feature extraction as well as classification performance; this process is optimized using the Migrating Walrus Algorithm for enhanced accuracy. The proposed approach obtained 99.9% accuracy as well as 99.8% precision for the IQ-OTH/NCCD database and 99.8% accuracy and 99.7% precision for the LIDC-IDRI database. These results show that the model works better than existing methods, suggesting a major advancement in LC diagnosis.