Hiding Patient Data in Medical Images in DICOM Metadata

Abstract

Keeping medical images and patient records in separate systems creates a huge security hole. Central databases are basically magnets for hackers. And if technicians strip out metadata to share a scan safely, doctors lose the clinical context they need. We fixed this structural flaw by locking patient data directly inside the DICOM file. We never touch the image pixels. Even minor pixel changes ruin a scan's diagnostic value. Metadata steganography bypasses that issue entirely. The security setup operates in layers. Patient text undergoes AES-256 encryption. We apply DNA cryptographic mapping next, and then hide the ciphertext inside reserved DICOM tags. The medical scan stays completely intact. During testing, the implementation hit an infinite Peak Signal-to-Noise Ratio (PSNR). That means zero visual distortion. Data extraction works perfectly. We also built in a SHA-256 hash check to catch any tampering instantly. Tying clinical identity straight to the imaging file lets the system ignore vulnerable central storage. It creates a highly secure, self-contained medical record.