Scholarlink

Class imbalance presents a significant challenge in Computational Pathology, particularly in classifying rare diseases such as renal amyloidosis. This paper investigates the effectiveness of Few-Shot Learning (FSL), specifically through prototypical networks, alongside conventional methods to enhance the automatic classification of renal glomeruli from biopsy images. A novel multi-stain dataset is introduced, comprising 11,674 annotated images across nine glomerular lesion classes, including amyloidosis, stained with four different dyes. The study compared baseline CNN models with FSL approaches, both with and without Cost-Sensitive Learning (CSL). The FSL-CSL-Ensemble achieved the highest F1-Score of 93.8%, surpassing the performance of related studies that addressed datasets with less severe imbalance ratios. This study underscores the potential of FSL in classifying renal amyloidosis, especially when combined with CSL, and suggests the possibility of eliminating the need for Congo red staining, the current gold standard for diagnosis. The findings highlight the necessity of developing innovative approaches like FSL to improve outcomes in medical image analysis, where data scarcity is prevalent.

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

Addressing Class Imbalance in Renal Amyloidosis Classification: A Comparative Study of Few-Shot Learning and Conventional Machine Learning Techniques

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Alexsandro Silva Santos,
  • Luciano Rebouças de Oliveira,
  • Washington Luis Conrado dos Santos,
  • Angelo Amancio Duarte

摘要

Class imbalance presents a significant challenge in Computational Pathology, particularly in classifying rare diseases such as renal amyloidosis. This paper investigates the effectiveness of Few-Shot Learning (FSL), specifically through prototypical networks, alongside conventional methods to enhance the automatic classification of renal glomeruli from biopsy images. A novel multi-stain dataset is introduced, comprising 11,674 annotated images across nine glomerular lesion classes, including amyloidosis, stained with four different dyes. The study compared baseline CNN models with FSL approaches, both with and without Cost-Sensitive Learning (CSL). The FSL-CSL-Ensemble achieved the highest F1-Score of 93.8%, surpassing the performance of related studies that addressed datasets with less severe imbalance ratios. This study underscores the potential of FSL in classifying renal amyloidosis, especially when combined with CSL, and suggests the possibility of eliminating the need for Congo red staining, the current gold standard for diagnosis. The findings highlight the necessity of developing innovative approaches like FSL to improve outcomes in medical image analysis, where data scarcity is prevalent.