Creating a Secure Data Sharing Network for Disease Identification Using CT Images and RHC-Based Encryption Scheme

This study integrates a dilated multiscale convolutional network with an encryption strategy based on Reversible Hidden Communication (RHC) to present a novel framework for safe data sharing and illness diagnosis using CT images. The system meets the vital requirement for safe data exchange in healthcare by guaranteeing the security, integrity, and privacy of medical data throughout transmission. Sensitive CT scans may be encrypted and decrypted reversibly and losslessly using RHC-based encryption, which makes it appropriate for sending massive amounts of medical data over unsecure networks. We create a dilated multiscale network for illness identification, using dilated convolutions to extract fine-grained information from the CT images at several scales. This network maintains spatial resolution while broadening the receptive field, improving the detection accuracy of conditions like tumours and lesions. The framework is designed to offer end-to-end security, enabling the safe exchange of encrypted medical pictures, their decryption, and subsequent analysis for the purpose of detecting diseases. This method guarantees high detection accuracy while striking a balance between security and computational economy. The suggested concept seeks to improve data privacy, expedite medical diagnoses, and advance the development of safer and more efficient remote healthcare and telemedicine systems. Future research will concentrate on enhancing the detection model's generalisation over a variety of datasets and speeding up encryption. Proposed method ShuffleNetV3 gave good result like 95% against all the other methods.