Scholarlink

<p>Deep learning achieves remarkable diagnostic accuracy in medical imaging, yet clinical adoption remains limited due to the black-box nature of neural networks. This paper introduces NS-MedNet, a hybrid neuro-symbolic framework integrating neural perception with symbolic medical reasoning for explainable image interpretation. The architecture combines a hybrid CNN-Transformer for feature extraction with a symbolic reasoning engine grounded in medical knowledge bases, producing diagnoses with clinically meaningful explanations. Experiments across six imaging modalities demonstrate 94.2% diagnostic accuracy with explanations rated 4.8/5.0 by clinical experts, significantly outperforming existing methods.</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

Hybrid Neuro-Symbolic Deep Learning Framework for Explainable Medical Image Interpretation

  • 期刊论文

  • 发表日期：

    2026年5月27日

查看全文
  • Priyanka Choudhary

摘要

Deep learning achieves remarkable diagnostic accuracy in medical imaging, yet clinical adoption remains limited due to the black-box nature of neural networks. This paper introduces NS-MedNet, a hybrid neuro-symbolic framework integrating neural perception with symbolic medical reasoning for explainable image interpretation. The architecture combines a hybrid CNN-Transformer for feature extraction with a symbolic reasoning engine grounded in medical knowledge bases, producing diagnoses with clinically meaningful explanations. Experiments across six imaging modalities demonstrate 94.2% diagnostic accuracy with explanations rated 4.8/5.0 by clinical experts, significantly outperforming existing methods.