Scholarlink

<p>This study presents an incremental learning framework to enhance the generalization and robustness of transformer-based deep learning models for segmenting skin cancer and related tissue structures. While deep learning models often perform well on data distributions similar to their training sets, their accuracy typically degrades when exposed to novel scenarios–limiting their clinical utility in skin cancer diagnosis. To address this, we propose a biologically inspired incremental learning strategy tailored for skin cancer classification and segmentation, allowing the model to incorporate new data progressively while reducing catastrophic forgetting. Our approach integrates multiple loss functions to preserve existing knowledge while adapting to additional magnification levels. Experimental results on the in-distribution test set demonstrate consistent performance improvements: achieving 89.05% accuracy with 10<InlineEquation ID="IEq1"> <EquationSource Format="TEX">\(\times\)</EquationSource> </InlineEquation> magnification, 92.68% with 10<InlineEquation ID="IEq2"> <EquationSource Format="TEX">\(\times\)</EquationSource> </InlineEquation> and 5<InlineEquation ID="IEq3"> <EquationSource Format="TEX">\(\times\)</EquationSource> </InlineEquation> combined, and 95.53% when incorporating 10<InlineEquation ID="IEq4"> <EquationSource Format="TEX">\(\times\)</EquationSource> </InlineEquation>, 5<InlineEquation ID="IEq5"> <EquationSource Format="TEX">\(\times\)</EquationSource> </InlineEquation>, and 2<InlineEquation ID="IEq6"> <EquationSource Format="TEX">\(\times\)</EquationSource> </InlineEquation> magnifications. These findings highlight the potential of our method to improve the adaptability and reliability of deep learning systems for empirical generalization in skin cancer classification tasks.</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

Incremental learning approach for semantic segmentation of skin histology images

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年3月6日

查看全文
  • Sana Fatima,
  • Anum Abdul Salam,
  • Muhammad Usman Akram,
  • Ibrahim A. Hameed,
  • Saad B. Ahmed

摘要

This study presents an incremental learning framework to enhance the generalization and robustness of transformer-based deep learning models for segmenting skin cancer and related tissue structures. While deep learning models often perform well on data distributions similar to their training sets, their accuracy typically degrades when exposed to novel scenarios–limiting their clinical utility in skin cancer diagnosis. To address this, we propose a biologically inspired incremental learning strategy tailored for skin cancer classification and segmentation, allowing the model to incorporate new data progressively while reducing catastrophic forgetting. Our approach integrates multiple loss functions to preserve existing knowledge while adapting to additional magnification levels. Experimental results on the in-distribution test set demonstrate consistent performance improvements: achieving 89.05% accuracy with 10 \(\times\) magnification, 92.68% with 10 \(\times\) and 5 \(\times\) combined, and 95.53% when incorporating 10 \(\times\) , 5 \(\times\) , and 2 \(\times\) magnifications. These findings highlight the potential of our method to improve the adaptability and reliability of deep learning systems for empirical generalization in skin cancer classification tasks.