Discrete Cosine Transform-based Data Compression of Bioimpedance with Embedded Hardware

This study explores the application of the Discrete Cosine Transform (DCT) to compress multi-frequency bioimpedance datasets that are well represented by the Cole-impedance model. The motivating goal is to reduce the data storage and wireless transmission requirements of these datasets to support portable and wearable systems that utilize this type of data. To evaluate the compression, DCT was applied to both synthetic and experimentally obtained datasets using MATLAB and an embedded-hardware environment (Arduino IDE and XIAO BLE board). Using this approach, compression ratios exceeding 98% were obtained for noiseless datasets, and up to 95% for noisy datasets, and 93% for experimental datasets. This validates the feasibility of using this DCT-based compression method for bioimpedance data to advance portable and wearable systems. Future efforts will focus on improving noise resilience and optimizing embedded hardware implementations.