Scholarlink

<p>Cardiac diseases are a significant threat to the humans in recent days. Predicting and diagnosing cardiac conditions is challenging. It requires detailed analysis of clinical data by medical professionals. However, most existing encryption techniques suffer from limitations in balancing robust security with computational efficiency. To overcome this challenge, a novel SCAM-MODEL is proposed for cardiac image encryption that secures data transmission and classifies cardiac diseases. The input images from the Sunnybrook dataset are scrambled and encrypted using a robust and highly secure Horsy Chess algorithm. The original image is recovered by applying the inverse of the encryption key movements during decryption. The decrypted image is pre-processing using a Gaussian Adaptive Bilateral filter to enhance quality while preserving edges and reducing noise. Capsule Network (CapsNet) is used to classify into Myocardial, Endocardial, Pericardial, Congenital, and Paediatric heart disease from the cardiac MRI image. The SCAM-MODEL achieves an NPCR of 99.46%, UACI of 33.44%, and IE of 7.44 bpp, and overall accuracy of 99.79% for Cardiac disease classification. The proposed approach improved the total accuracy by 0.44%, 0.72%, and 0.52% compared to SHA-256, GSAPSO-MQC, and ResNet-50, respectively.</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

SCAM-MODEL: Secure Cardiac MRI Image Encryption via Lightweight Pixel-Permutation-Based Horsy Chess Model

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年1月30日

查看全文
  • K. Gopalakrishnan,
  • A. Ahilan,
  • P. Deepa,
  • N. Muthukumaran

摘要

Cardiac diseases are a significant threat to the humans in recent days. Predicting and diagnosing cardiac conditions is challenging. It requires detailed analysis of clinical data by medical professionals. However, most existing encryption techniques suffer from limitations in balancing robust security with computational efficiency. To overcome this challenge, a novel SCAM-MODEL is proposed for cardiac image encryption that secures data transmission and classifies cardiac diseases. The input images from the Sunnybrook dataset are scrambled and encrypted using a robust and highly secure Horsy Chess algorithm. The original image is recovered by applying the inverse of the encryption key movements during decryption. The decrypted image is pre-processing using a Gaussian Adaptive Bilateral filter to enhance quality while preserving edges and reducing noise. Capsule Network (CapsNet) is used to classify into Myocardial, Endocardial, Pericardial, Congenital, and Paediatric heart disease from the cardiac MRI image. The SCAM-MODEL achieves an NPCR of 99.46%, UACI of 33.44%, and IE of 7.44 bpp, and overall accuracy of 99.79% for Cardiac disease classification. The proposed approach improved the total accuracy by 0.44%, 0.72%, and 0.52% compared to SHA-256, GSAPSO-MQC, and ResNet-50, respectively.