A key technology for compressing biometric signals developed by South Korean researchers has been adopted as an international standard. This marks a significant step in filling the gap in international standards for biometric signal compression, which has been largely absent compared to medical imaging.
The Electronics and Telecommunications Research Institute (ETRI) announced on July 29 that its proprietary biometric signal compression technology has been officially incorporated into the international standard for biometric and general waveform compression, jointly promoted by the International Telecommunication Union's Telecommunication Standardization Sector (ITU-T) and the Moving Picture Experts Group (MPEG). The standard was approved simultaneously as ITU-T Recommendation T.261 and MPEG-D Part 8 during a standardization meeting in July.
ETRI not only contributed two core compression technologies to the international standard but also took on the role of editor for the standard document, leading the international standardization effort. The biometric and general waveform compression international standard (H.BWC) efficiently compresses biometric signals such as electroencephalograms (EEG), electrocardiograms (ECG), and electromyograms (EMG) while preserving the original data.
While medical imaging has widely adopted the DICOM international standard, there has been no comparable standard for biometric signal compression. Biometric signals generate large data volumes, and even minor errors can have critical impacts on medical diagnoses, making lossless data reduction technology essential.
ETRI's standard incorporates two key compression technologies: pre-predictive coding and electrode placement-based channel coding. The pre-predictive coding technology predicts the next value based on the previous signals and encodes only the difference between the predicted and actual values, thereby reducing the amount of information that needs to be compressed and enhancing efficiency. The electrode placement-based channel coding technology addresses inefficiencies in multi-channel signal measurement during EEG tests, where identical information may be redundantly recorded across different channels. By identifying redundancy based on the connection information between electrodes, this technology prevents unnecessary duplicate data storage and efficiently reconstructs signals with minimal additional computational burden, achieving superior compression efficiency compared to existing methods.
Both technologies underwent rigorous international joint validation before being adopted as standard technologies, and the research team has also secured two international standard patents related to this work.
The introduction of this technology allows for a significant reduction in file sizes while maintaining the quality of original biometric signal data, enabling medical institutions to efficiently store and transmit large volumes of medical data. It is expected to be utilized in the converging healthcare industry, which requires high-quality biometric signal processing for telemedicine, wearable healthcare, and AI-based precision medical diagnostics. Additionally, it can be applied to the compression of general waveform data, such as industrial sensor data collection and vibration analysis, and this lossless compression technology has also been adopted for MPEG's audio compression standard for machines (ACoM), expanding its range of applications.
“This achievement signifies that our technology has been officially recognized as a core component of international standards, going beyond merely proposing a technology,” said Kang Jeong-won, head of ETRI's Media Coding Research Lab. “We will continue to advance compression technologies in the fields of biometric signals and immersive media and lead international standardization efforts.”
* This article has been translated by AI.
Copyright ⓒ Aju Press All rights reserved.