Scholarlink

Cancer has always garnered significant attention due to its prevalence and the impact it has on many individuals. Recent advancements in artificial intelligence (AI) have led to numerous research initiatives that effectively assist doctors in diagnosis and treatment, thus enhancing patients’ chances of recovery. Mammography remains one of the most commonly used screening methods today. In this paper, we propose a detection and classification method for mammography that integrates a fuzzy inference system (FIS) with explainable artificial intelligence (XAI) to enhance classification efficiency and improve the interpretation of outputs from machine learning models. We conduct experiments across various scenarios to assess the model’s practical applicability. The experimental results demonstrate an accuracy exceeding 94.64% and an IoU index of 0.899, confirming the proposed method’s effectiveness. These findings lay a crucial foundation for our ongoing development of fuzzy learning techniques aimed at tumor classification and identification in mammography.

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

An Explainable AI-Based Breast Cancer Classification Combining Rule Sets and Deep Neural Networks

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Ho-Dat Tran,
  • Thuong-Cang Phan,
  • Vinh-Phong Nguyen,
  • Anh-Cang Phan

摘要

Cancer has always garnered significant attention due to its prevalence and the impact it has on many individuals. Recent advancements in artificial intelligence (AI) have led to numerous research initiatives that effectively assist doctors in diagnosis and treatment, thus enhancing patients’ chances of recovery. Mammography remains one of the most commonly used screening methods today. In this paper, we propose a detection and classification method for mammography that integrates a fuzzy inference system (FIS) with explainable artificial intelligence (XAI) to enhance classification efficiency and improve the interpretation of outputs from machine learning models. We conduct experiments across various scenarios to assess the model’s practical applicability. The experimental results demonstrate an accuracy exceeding 94.64% and an IoU index of 0.899, confirming the proposed method’s effectiveness. These findings lay a crucial foundation for our ongoing development of fuzzy learning techniques aimed at tumor classification and identification in mammography.