Scholarlink

Skin cancer is still among the most prevalent global health issues, yielding high mortality rates and a large number of affected individuals. The use of effective and valid screening tools to detect skin cancer early is a crucial need to enhance treatment success. Ultraviolet (UV) radiation exposure is a significant agent behind the increase in both melanoma and non-melanoma skin cancers, thereby adding an extra burden to healthcare systems. The primary challenge lies in developing solutions that are both cost-effective and highly accurate for early detection. Automated systems are therefore necessary with the ability to read dermoscopic images. This research proposes an automatic solution for identifying skin cancer with specific focus given to detecting the non-melanoma cases employing MobileNet lightweight, compact Convolutional Neural Network (CNN)-optimized model tailored to run optimally on constrained hardware.

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

Detection and Classification of Non-melanoma Skin Cancer Using Deep Learning Model

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Sreekara Karthik Gundlapalli,
  • Gnanesh Vaka,
  • P. Saranya

摘要

Skin cancer is still among the most prevalent global health issues, yielding high mortality rates and a large number of affected individuals. The use of effective and valid screening tools to detect skin cancer early is a crucial need to enhance treatment success. Ultraviolet (UV) radiation exposure is a significant agent behind the increase in both melanoma and non-melanoma skin cancers, thereby adding an extra burden to healthcare systems. The primary challenge lies in developing solutions that are both cost-effective and highly accurate for early detection. Automated systems are therefore necessary with the ability to read dermoscopic images. This research proposes an automatic solution for identifying skin cancer with specific focus given to detecting the non-melanoma cases employing MobileNet lightweight, compact Convolutional Neural Network (CNN)-optimized model tailored to run optimally on constrained hardware.