Enhanced Model for Medical Image Data Security Using Machine Learning

In the contemporary landscape of healthcare, the burgeoning development of Internet-based health services harnesses advancements in data communication technologies, facilitating the widespread sharing of vast medical data among professionals worldwide. Recent strides in Information Technology (IT) have significantly contributed to effective medical data management. This study addresses challenges associated with the secure transmission of medical images, highlighting potential risks such as inappropriate diagnosis reports, authentication issues, and compromised data privacy due to variations in transmitted medical data. The Enhanced Model for Medical Image Data Security (EM-MIDS) is introduced, utilizing machine learning techniques to fortify the security of medical image data. K-Means Clustering (KMC) is employed to cluster similar pixels within input medical images, forming the basis for subsequent image encryption and decryption processes. Chaotic Map and channel-based data security methodologies are integrated into the encryption and decryption procedures. For accurate classification, the model employs Random Forest (RF), while Support Vector Machine (SVM) assesses overall data security. Experimental evaluation, using benchmark medical images, comprehensively assesses EM-MIDS. Performance evaluation criteria are applied to scrutinize the model’s efficiency. Comparative analyses with existing models gauge the superiority of the proposed approach. This research contributes to establishing a robust and secure framework for medical image data transmission, addressing concerns of accuracy, privacy, and data integrity. The integration of machine learning techniques enhances the effectiveness of EM-MIDS, positioning it as a promising solution for securing medical data in the digital healthcare landscape.