Efficient EEG Data Compression Using Cellular Automata: A Rule 90 and Rule 150-Based Approach

Abstract This paper presents an innovative Cellular Automata (CA)-based framework for efficient compression of electroencephalogram (EEG) data, leveraging Rule 90 and Rule 150. These specific CA rules are selected due to their computa- tional efficiency and their ability to preserve critical seizure-related patterns within EEG signals. Rule 90, a linear rule, maintains structural integrity and preserves essential bit transition patterns. In contrast, Rule 150, a chaotic rule, enhances edge detection and local variations, ensuring that diagnostically rel- evant features remain intact. The proposed approach efficiently compresses 28 input samples into just 4 output samples, achieving an 85% reduction in storage requirements and significantly minimizing memory utilization. This optimization enhances its suitability for large-scale, real-time EEG signal processing applica- tions. The proposed CA-based data compression framework has been efficiently implemented on an Xilinx Zynq 7000 FPGA, demonstrating a total power con- sumption of 219mW, with 111mW allocated to dynamic power and 108mW to static power. The results indicate that the proposed CA-based compression methodology offers a highly efficient and hardware-optimized solution for EEG data processing, making it well-suited for low-power, embedded, and real-time biomedical applications.