Scholarlink

<p>Pharyngitis and tonsillitis are common throat inflammations, with pharyngitis affecting the pharyngeal mucosa and tonsillitis targeting the tonsils. Both conditions present overlapping symptoms, complicating diagnosis and treatment. Oral cancer, a serious malignancy in the mouth or throat, poses significant health risks, often linked to tobacco, alcohol, or HPV, and requires early detection for better outcome. The research concerning the classification of pharyngitis, tonsillitis, and oral cancer encountered challenges such as model overfitting, suboptimal classifier training, and unreliable accuracy metrics, which collectively limited the robustness and generalizability of the existing approaches. Thus, to tackle the challenges of the existing method, Bidirectional Block Self-attention (BiBloSAN) and Self-adaptive pooling Enhanced attention (SPEM)-enabled Quantum Convolutional Neural Network ((BS)<sup>2</sup>Att-QCNN) is proposed in the research. The combination of attention mechanism into the QCNN enables accurate classification. Further, the segmentation in the research is performed with the optimization called Spheniscidae Spiral Huddling Algorithm (S2HA) that aids in obtaining the efficient segmentation of the inputs, which aids the overall research to obtain high classification accuracy. The overall efficiency of the research is evaluated on aggregated bench mark data sets with metrics such as precision, sensitivity, accuracy, recall, f1-score, and specificity with attained values of 95.07%, 96.22%, 95.78%, 96.22%, 95.64%, and 95.36% 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

Speech organ disease classification using bidirectional block self-adaptive attention mechanism enabled quantum convolutional neural network

  • 期刊论文

  • 发表日期：

    2025年9月14日

查看全文
  • M. Swathi,
  • Rajeshkannan Regunathan

摘要

Pharyngitis and tonsillitis are common throat inflammations, with pharyngitis affecting the pharyngeal mucosa and tonsillitis targeting the tonsils. Both conditions present overlapping symptoms, complicating diagnosis and treatment. Oral cancer, a serious malignancy in the mouth or throat, poses significant health risks, often linked to tobacco, alcohol, or HPV, and requires early detection for better outcome. The research concerning the classification of pharyngitis, tonsillitis, and oral cancer encountered challenges such as model overfitting, suboptimal classifier training, and unreliable accuracy metrics, which collectively limited the robustness and generalizability of the existing approaches. Thus, to tackle the challenges of the existing method, Bidirectional Block Self-attention (BiBloSAN) and Self-adaptive pooling Enhanced attention (SPEM)-enabled Quantum Convolutional Neural Network ((BS)2Att-QCNN) is proposed in the research. The combination of attention mechanism into the QCNN enables accurate classification. Further, the segmentation in the research is performed with the optimization called Spheniscidae Spiral Huddling Algorithm (S2HA) that aids in obtaining the efficient segmentation of the inputs, which aids the overall research to obtain high classification accuracy. The overall efficiency of the research is evaluated on aggregated bench mark data sets with metrics such as precision, sensitivity, accuracy, recall, f1-score, and specificity with attained values of 95.07%, 96.22%, 95.78%, 96.22%, 95.64%, and 95.36% respectively.