Scholarlink

<p>Out of many irreversible eye diseases worldwide, glaucoma stands out as the primary disease of blindness. Early detection and diagnosis play a crucial role in the treatment of glaucoma. However, its early detection is quite challenging, since glaucoma is asymptomatic in its nature. Thus, manual screening is tedious and error prone. Fortunately, in recent years, computer-vision plays a major role in auto screening of medical images. This enables us to have a quick and qualitative diagnosis to save the patient. In this proposal, a novel deep learning-based glaucoma classification approach using retinal images has been introduced. This approach begins with the pre-processing of input images to enhance their texture quality. Then the image patterns are recognized by the proposed deep convolutional networks using various experimental combinations. This methodology utilizes retinal datasets such as DRISHTI, ORIGA, and ACRIMA. This approach aims to reduce the likelihood of early misdiagnosis and missed diagnoses of glaucoma, enabling doctors to make more accurate judgments, design personalized treatment plans, and prevent further impairment of visual function in patients.</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 deep retinal vision network for glaucoma classification

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年2月25日

查看全文
  • Krishna Santosh Naidana,
  • Madhu Hasitha Manne,
  • Hema Yalavarthi

摘要

Out of many irreversible eye diseases worldwide, glaucoma stands out as the primary disease of blindness. Early detection and diagnosis play a crucial role in the treatment of glaucoma. However, its early detection is quite challenging, since glaucoma is asymptomatic in its nature. Thus, manual screening is tedious and error prone. Fortunately, in recent years, computer-vision plays a major role in auto screening of medical images. This enables us to have a quick and qualitative diagnosis to save the patient. In this proposal, a novel deep learning-based glaucoma classification approach using retinal images has been introduced. This approach begins with the pre-processing of input images to enhance their texture quality. Then the image patterns are recognized by the proposed deep convolutional networks using various experimental combinations. This methodology utilizes retinal datasets such as DRISHTI, ORIGA, and ACRIMA. This approach aims to reduce the likelihood of early misdiagnosis and missed diagnoses of glaucoma, enabling doctors to make more accurate judgments, design personalized treatment plans, and prevent further impairment of visual function in patients.