Organic Neuromorphic Circuits for Real-Time Biosignal Applications

Organic neuromorphic circuits offer new opportunities for low-power, flexible electronics capable of real-time inference at the edge. In this work, we present a neuromorphic system based on organic thin-film transistors (OTFTs) that performs biosignal classification in a Bayesian framework. The spiking behavior of artificial OTFT neuron circuits are first characterized, showing how their dynamics can be tuned through circuit parameters. We then demonstrate how the circuits can infer information from real-world electroencephalography (EEG) data. When applied to epilepsy-related signal patterns, the system achieves excellent classification performance, while maintaining ultra-low power operation. These results highlight the potential of OTFT-based neuromorphic architectures for embedded medical diagnostics.