Scholarlink

Healthcare practitioners use X-rays as a vital diagnostic tool for a variety of illnesses. But it’s vital to remember that prompt and precise diagnosis is essential to good patient care and treatment. Although chest X-rays can yield extremely accurate anatomical information, manual picture understanding can be laborious and error-prone, potentially resulting in delays or misdiagnosis. In an effort to improve the precision and effectiveness of data-driven medical diagnosis, the field of machine learning-based X-ray image identification and recognition is expanding quickly. Classifying X-rays is difficult because to the low resolution, noise in the images, and lack of comparison. The machine learning-based X-ray image identification and recognition process is demonstrated in this study using a CNN architecture-based ResNet50 model. The data set for this study was created using two picture pre-processing techniques: morphological image processing and the Hough transform. The primary goal that is being investigated outlined is to use the ResNet50 architecture to create a system that can recognize lung cancer in photos. The ResNet50 model was trained using this data collection to detect lung cancer. Evaluation criteria were used to assess the precision, sensitiveness, and specificity of the model. The results showed a general accurateness of \(96.87\%,\) a sensitivity of \(98.21,\) and a specificity of \(95.63,\) all of which are superior than the current methods. The results indicated that X-ray pictures might potentially be efficiently classified using the ResNet50 model.

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

Assessment on Convolutional Neural Networks for Machine Learning-Based X-Ray Image Detection and Categorization

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • X. S. Asha Shiny,
  • Singanamala Priyanka,
  • M. Srilekha,
  • J. Prasanna Babu,
  • Komal Parashar,
  • K. Venkateswara Rao

摘要

Healthcare practitioners use X-rays as a vital diagnostic tool for a variety of illnesses. But it’s vital to remember that prompt and precise diagnosis is essential to good patient care and treatment. Although chest X-rays can yield extremely accurate anatomical information, manual picture understanding can be laborious and error-prone, potentially resulting in delays or misdiagnosis. In an effort to improve the precision and effectiveness of data-driven medical diagnosis, the field of machine learning-based X-ray image identification and recognition is expanding quickly. Classifying X-rays is difficult because to the low resolution, noise in the images, and lack of comparison. The machine learning-based X-ray image identification and recognition process is demonstrated in this study using a CNN architecture-based ResNet50 model. The data set for this study was created using two picture pre-processing techniques: morphological image processing and the Hough transform. The primary goal that is being investigated outlined is to use the ResNet50 architecture to create a system that can recognize lung cancer in photos. The ResNet50 model was trained using this data collection to detect lung cancer. Evaluation criteria were used to assess the precision, sensitiveness, and specificity of the model. The results showed a general accurateness of \(96.87\%,\) a sensitivity of \(98.21,\) and a specificity of \(95.63,\) all of which are superior than the current methods. The results indicated that X-ray pictures might potentially be efficiently classified using the ResNet50 model.