Scholarlink

<p>Breast cancer (BC) is one of the most prevalent malignancies worldwide, affecting millions of women. It is also a major cause of cancer-related mortality among women. Several deep learning (DL) models have been developed in recent years to assist in breast cancer diagnosis. However, many existing models still exhibit limited accuracy, which can lead to misdiagnosis. In order to overcome these constraints and enhance diagnostic accuracy, this research proposed&#xa0;a unique method called Self-Guided Quantum Generative Adversarial Network with Puma Optimizer (SGQGAN-PO). The INBreast and DDSM datasets were used to evaluate the proposed methodology. The Square-Root Sage-Husa Adaptive Robust Kalman Filter (SR-SHARKF) was applied to enhance contrast and reduce noise in mammography images. The segmentation model U-Net with Diffusion Kernel Attention Network (U-Net + DKAN) compresses images, captures fine characteristics, and reconstructs segmentation masks using an encoder-decoder structure and skip connections. A Spike-driven Transformer (SDT), based on the Leaky Integrate-and-Fire neuron model, provides efficient, event-driven feature extraction. Classification is then performed using an SGQGAN to detect breast cancer. The SGQGAN hyperparameters are optimized using the PO. The accuracy concerning 99.5% in the dataset for DDSM and 99.8% for the INBreast dataset was obtained by evaluating the (SGQGAN -PO) model's efficiency utilizing different datasets. Such results show that the model is more powerful than the existing approaches, so it can be further advanced in the field.</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

Enhanced Breast Cancer Detection in Mammographic Images Using Self-Guided Quantum Generative Adversarial Network with Puma Optimizer

  • 期刊论文

  • 发表日期：

    2026年6月3日

查看全文
  • P. B. V. Raja Rao,
  • J. Savija,
  • Elangovan Muniyandy,
  • Hessan Mohammad

摘要

Breast cancer (BC) is one of the most prevalent malignancies worldwide, affecting millions of women. It is also a major cause of cancer-related mortality among women. Several deep learning (DL) models have been developed in recent years to assist in breast cancer diagnosis. However, many existing models still exhibit limited accuracy, which can lead to misdiagnosis. In order to overcome these constraints and enhance diagnostic accuracy, this research proposed a unique method called Self-Guided Quantum Generative Adversarial Network with Puma Optimizer (SGQGAN-PO). The INBreast and DDSM datasets were used to evaluate the proposed methodology. The Square-Root Sage-Husa Adaptive Robust Kalman Filter (SR-SHARKF) was applied to enhance contrast and reduce noise in mammography images. The segmentation model U-Net with Diffusion Kernel Attention Network (U-Net + DKAN) compresses images, captures fine characteristics, and reconstructs segmentation masks using an encoder-decoder structure and skip connections. A Spike-driven Transformer (SDT), based on the Leaky Integrate-and-Fire neuron model, provides efficient, event-driven feature extraction. Classification is then performed using an SGQGAN to detect breast cancer. The SGQGAN hyperparameters are optimized using the PO. The accuracy concerning 99.5% in the dataset for DDSM and 99.8% for the INBreast dataset was obtained by evaluating the (SGQGAN -PO) model's efficiency utilizing different datasets. Such results show that the model is more powerful than the existing approaches, so it can be further advanced in the field.