Design and Evaluation of a Compact EEG Acquisition Platform for Wearable Applications

Unlike traditional, bulky electroencephalography (EEG) systems used primarily in clinical or laboratory settings, wearable EEG devices enable flexible, non-invasive brain signal monitoring during daily activities [1]. With recent advancements in artificial intelligence (AI) and low-power wireless technologies, these devices are becoming more effective for continuous, personalized monitoring and management of neurological conditions [2]. Epilepsy, in particular, can benefit from wearable EEG systems for real-time seizure prediction and timely intervention [3].<br/>This work presents the design and evaluation of a compact, low-power EEG acquisition platform optimized for wearable applications. The prototype (35 × 25 mm) (see Fig. 1 (1)) integrates a memory chip, a Bluetooth low energy (BLE)-enabled microcontroller, an inertial measurement unit (IMU), and a multi-channel EEG Analog Front End (AFE). Power consumption, measured at 4.2 V using a Keithley 2450 source meter, averaged 36.8 mA, supporting 9.5 hours of continuous operation with a 350 mAh LiPo battery (Fig. 1 (3)). The 8 Gbit memory enables continuous operation for over 30.5 hours.<br/>Signal accuracy was assessed using a waveform generator and attenuator to provide sinusoidal inputs (10 μVpp–40 mVpp at 5 Hz, except 25 mVpp at 10 Hz; Fig. 1(2)). Results are averages from repeated tests on 5-second signals at a PGA of 1 and a 250 Hz sampling rate. The platform accurately captured low-amplitude signals, matching the input waveforms, as shown in the statistical analysis (Fig. 1(4)).<br/>These results demonstrate the platform’s suitability for real-world EEG monitoring, balancing compactness, efficiency, and signal accuracy. Future work will deploy on-edge Machine Learning (ML)-based models for enhanced real-time signal processing.