Low-Complexity CNN-Based Classification of Electroneurographic Signals

TL;DR

This work tackles real-time classification of electroneurographic signals under stringent resource constraints typical of implantable peripheral nerve interfaces. It introduces MobilESCAPE-Net, a low-complexity CNN built upon spatiotemporal ENG signatures, depthwise separable convolutions, and Global Average Pooling to substantially reduce parameters and FLOPs while maintaining or improving accuracy and macro F1-score relative to ESCAPE-Net. The paper demonstrates a ~99.9% reduction in trainable parameters and ~92.5% reduction in FLOPs, with MobilESCAPE-Net achieving comparable performance across multiple rats and stimuli, and robustness under varying signal-to-noise ratios. This enables faster, real-time ENG decoding suitable for deployment in resource-constrained ND&S systems and implantable devices.

Abstract

Peripheral nerve interfaces (PNIs) facilitate neural recording and stimulation for treating nerve injuries, but real-time classification of electroneurographic (ENG) signals remains challenging due to constraints on complexity and latency, particularly in implantable devices. This study introduces MobilESCAPE-Net, a lightweight architecture that reduces computational cost while maintaining and slightly improving classification performance. Compared to the state-of-the-art ESCAPE-Net, MobilESCAPE-Net achieves comparable accuracy and F1-score with significantly lower complexity, reducing trainable parameters by 99.9\% and floating point operations per second by 92.47\%, enabling faster inference and real-time processing. Its efficiency makes it well-suited for low-complexity ENG signal classification in resource-constrained environments such as implantable devices.

Figures (11)

• Figure 1: Cuff electrode model with example ENG signals recorded for dorsiflexion, plantarflexion, and pricking.
• Figure 3:
• Figure 4:
• Figure 5:
• Figure 7: MobilESCAPE-Net architecture.