What question did this study set out to answer?

The aim is to improve ECG compression methods to efficiently handle large data volumes in IoMT applications.

March 1, 2026

Ultra-High Quality ECG Compression for IoMT Application Using Temporal Convolutional Auto-Encoder with Improved RVQ

Key Result

The proposed temporal-convolutional auto-encoder with RVQ achieved 88× ECG compression ratio, QS of 42.7, PRD 2.36%, outperforming existing methods on MIT BIH data.

Key Points

The aim is to improve ECG compression methods to efficiently handle large data volumes in IoMT applications.
Developed a multi-granularity temporal-convolutional auto-encoder architecture
Integrated Residual Vector Quantization within the encoder-decoder pipeline
Introduced a codebook-projection layer for better codebook use and fidelity
Implemented periodicity-aware modeling to capture ECG dynamics
Achieved a compression ratio of 88×
Obtained a Quality Score of 42.7
Maintained Percentage Root Mean Difference at 2.36%
Percentages Root Mean Difference Normalized reported at 17.56%
Demonstrated generalization via zero shot evaluation on another dataset

Structured PICO

Population

ECG records from the MIT BIH Arrhythmia Database and PhysioNet-2017 dataset

Intervention

Multi-granularity temporal-convolutional auto-encoder with Residual Vector Quantization (RVQ)

Comparator

Existing ECG compression techniques

Outcome

Compression ratio, Quality Score (QS), Percentage Root Mean Difference (PRD), and Percentage Root mean Difference Normalized (PRDN)

The proposed temporal-convolutional auto-encoder with RVQ achieves high ECG compression ratios (88x) while maintaining high reconstruction fidelity, facilitating prolonged ECG monitoring in IoMT applications.

Main Result

Absolute Event Rate: 0% vs 0%

Abstract

Prolonged Electrocardiogram (ECG) monitoring through the Internet of Medical Things (IoMT) is vital for cardiac diagnosis yet generates prohibitive data volumes, posing significant challenges to storage and transmission. However, conventional ECG compression paradigms have plateaued, failing to push compression ratios higher under stringent fidelity constraints. To address these limitations, we propose an end-to-end architecture that synergizes a multi-granularity temporal-convolutional auto-encoder with Residual Vector Quantization (RVQ). The design introduces three complementary components: (1) RVQ integrated in the encoder-decoder pipeline to boost compression ratios; (2) a codebook-projection layer that increases codebook utilization and reconstruction fidelity; (3) periodicity-aware modeling that captures intrinsic ECG dynamics and further suppresses distortion. Extensive experiments on the MIT BIH Arrhythmia Database show that the proposed method attains a compression ratio of 88× and Quality Score (QS) of 42.7, while keeping the Percentage Root Mean Difference (PRD) at 2.36% and the Percentage Root mean Difference Normalized (PRDN) at 17.56%, clearly surpassing existing techniques. Generalization is further confirmed by zero shot evaluation on the PhysioNet-2017 dataset. Overall, this paper presents an effective end-to-end compression frame work, and experimental results corroborate its efficacy. The code andmodelareopen-sourced at the repository https: //github.com/Guan-Y/ECG-Codec.

Bookmark