Citation

S. Thamil Selvan, Thivaagar (2025) Hybrid compression-encryption models on biomedical signals for telemedicine and storage applications. Masters thesis, Multimedia University.

Full text not available from this repository.

Abstract

An electrocardiogram (ECG) is an electronic record of a patient’s heartbeat. ECG data is often stored in hospitals for further analysis and transmitted online from remote areas. Storing or transmitting enormous amounts of ECG signals to another device or online is an unendurable process without compressing the signals. The ECG data also needs to be protected from unauthorized access during the transmission, as it may hold sensitive information of the patient. In this study, we propose a hybrid method for ECG compression and encryption that combines long short-term memory (LSTM) as a predictor with Huffman encoding for compression, and Advanced Encryption Standard (AES) with Cipher Block Chaining (CBC) mode for encryption. The proposed method aims to achieve both size reduction and data security for ECG data. To the best of our knowledge, LSTM as a predictor for compression has not been explored by other researchers. For performance evaluation on quality and compression efficiency, various measures were used to test our proposed model, such as compression ratio (CR) and percentage root mean square difference (PRD). Encryption efficiency is evaluated by calculating encryption and decryption times. The results indicate that our hybrid method provides a higher compression ratio, stronger encryption, and better execution efficiency compared to existing methods. The evaluation was conducted on 48 records, and each about 30 minutes long from the MIT-BIH Arrhythmia Database. Encryption and decryption were completed in under 5 milliseconds per record. In this research, a CR of 2.78 can be achieved on average with the PRD of 0, which shows that our lossless compression method achieves a higher compression ratio and stronger encryption compared to existing methods, while maintaining the confidentiality of the original ECG data. In addition, our method offers a convenient and efficient way to protect and transmit sensitive ECG data, as it combines both techniques into a single process. It has the potential to improve the storage and transmission of ECG data in clinical settings. The current study is limited to analysis of previously recorded data from MIT-BIH database using compression based on LSTM and Huffman coding and encryption using AES-CBC only. Future studies may explore other compression and encryption techniques.

Item Type:	Thesis (Masters)
Additional Information:	Call No.: QA76.9.A25 T45 2025
Uncontrolled Keywords:	Data encryption (Computer science)
Subjects:	Q Science > QA Mathematics > QA71-90 Instruments and machines
Divisions:	Faculty of Computing and Informatics (FCI)
Depositing User:	Ms Nurul Iqtiani Ahmad
Date Deposited:	22 May 2026 08:18
Last Modified:	22 May 2026 08:18
URII:	http://shdl.mmu.edu.my/id/eprint/15905

Downloads

Edit (login required)