Scholarlink

<p>The challenge of extracting features from biomedical images with complex compositions is considered. A&#xa0;system for generating descriptions of biomedical images has been developed, which allows doctors and specialists to calculate the features of objects of study interest with subsequent analysis of the informativeness of the features. With the aim of analyzing the effectiveness of the system developed here, studies were conducted for two types of objects—bone marrow cells and pathological and normal computed tomography images. The experiments showed that in the case of bone marrow images containing nuclei, an accuracy of 78.52% was obtained using features based on the spatial adjacency and series length matrixes. For computed tomography images, an accuracy of 91.91% was obtained based on fractal and Fourier features.</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

A system for generating descriptions of images of medical objects

  • 期刊论文

  • 发表日期：

    2025年11月10日

查看全文
  • E. V. Polyakov,
  • E. V. Shuvalova,
  • V. V. Dmitrieva

摘要

The challenge of extracting features from biomedical images with complex compositions is considered. A system for generating descriptions of biomedical images has been developed, which allows doctors and specialists to calculate the features of objects of study interest with subsequent analysis of the informativeness of the features. With the aim of analyzing the effectiveness of the system developed here, studies were conducted for two types of objects—bone marrow cells and pathological and normal computed tomography images. The experiments showed that in the case of bone marrow images containing nuclei, an accuracy of 78.52% was obtained using features based on the spatial adjacency and series length matrixes. For computed tomography images, an accuracy of 91.91% was obtained based on fractal and Fourier features.