A Biologically Interfaced Evolvable Organic Pattern Classifier

Area of Science:

• Neuroscience
• Materials Science
• Electrical Engineering
• Biotechnology

Background:

• Future brain-computer interfaces (BCIs) necessitate localized, personalized signal processing integrated within biological tissues.
• Existing BCIs require advanced hardware capable of processing sensor data and translating it for biological interpretation.
• Organic electrochemical transistors (OECTs) offer potential for bio-integrated electronics but face challenges in stability and performance.

Purpose of the Study:

• To report the first hardware-based pattern classifier interfaced with a biological nerve.
• To develop an adaptive neural interface using evolvable organic electrochemical transistors (EOECTs).
• To demonstrate selective muscle stimulation based on processed neural input patterns.

Main Methods:

• Implementation of the Widrow-Hoff learning algorithm on an array of EOECTs for pattern classification.
• In situ modulation of EOECT channel conductance via electropolymerization for enhanced device characteristics.
• Interfacing the organic classifier with a biological nerve using an organic electrochemical spiking neuron for signal translation.

Main Results:

• Demonstrated low voltage operation, high reproducibility, and a two-orders-of-magnitude improvement in state retention for EOECT devices.
• Successfully translated classifier output into simulated action potentials using an organic electrochemical spiking neuron.
• Achieved selective muscle contraction stimulation based on the input pattern, validating the closed-loop system.

Conclusions:

• The developed organic pattern classifier represents a significant advancement in bio-integrated electronic systems.
• This EOECT-based system provides a robust platform for adaptive neural interfaces.
• The findings pave the way for developing sophisticated closed-loop therapeutic systems for neural applications.