Development of a cryptography model based on improved filtering, compression and encryption techniques for ECG signal processing

Electrocardiography is the process of producing an electrocardiogram (ECG) which is a convenient tool for identifying people with potential heart diseases which may need immediate referral to a hospital or emergency medical services in E-healthcare. The ECG signal remote monitoring application in the healthcare services face many challenges related to the real-time diagnosis. The noise cancellation of the ECG signal is critical for accurate extraction of useful heart data from ECG. Additionally, the continuous flow of signals may lead to a sheer increase in the volume of the data, the ECG data needs a large memory storage device. At the same time, security and privacy of the data is considered as a significant aspect of remote diagnosis medical application that further increases the volume of data sharing, including the risk factor. This research work proposed a model to combine approaches for ECG denoising, data encoding, and encryption. Further, improved ECG signal processing based on improved filtering, an adaptive lossless compression mechanism, and hybrid cryptography are proposed. For the denoising of the ECG signal, an enhanced and extended Kalman and adaptive Recursive Least Square (RLS) filtering have been used for signal filtering along with Discrete Wavelet Transform (DWT). The compression mechanism is performed using adaptive lossless compression based on Huffman encoding. Furthermore, to increase security, a cryptography mechanism has been employed using the Advanced Encryption Standard (AES) algorithm and Cipher Block Chaining (CBC) operation mode scheme with a 256-bit key. The Diffie-Hellman key exchange and Rivest Shamir Adleman (RSA) key generation algorithms have been used to authenticate the receiver, and key generation for encrypting and decrypting processes, respectively. Consequently, the main contributions of this research work include a high level of security, privacy, encoding with low error reconstruction along with reduced noise and processing time for the ECG signal in e-healthcare services. The proposed model is for denoising, assuring data security, and compression performance for ECG data storage and transmission on MIT-BIH and PTB Diagnostic ECG dataset. The experimental results show that the proposed system model is successfully the denoising, and secure storage and transmission of ECG data. Based on the simulation results show a decrease for SNR by SNRimp of 55 in dB, a significant improvement of 21.92 for MSE and good accuracy for PSNR and CC. Furthermore, the throughput average of CR is enhanced by 26.66 and 0.8416 for PRD compared with existing different compression schemes for the ECG signal. Finally, the proposed system model is utilized for high-level security against for various kinds of attacks such as denial-of-service (DoS), Distributed DoS, privacy attack, and Man-in-the-middle (MitM)