Neuromorphic Imaging Flow Cytometry combined with Adaptive Recurrent Spiking Neural Networks

TL;DR

This paper introduces a neuromorphic imaging flow cytometer that uses event-based CMOS cameras and adaptive spiking neural networks to classify particles with high accuracy and low latency, advancing biomedical imaging technology.

Contribution

The study demonstrates the integration of neuromorphic imaging with adaptive recurrent spiking neural networks for real-time particle classification in flow cytometry, achieving high accuracy and efficiency.

Findings

Recurrent spiking neural networks achieved 98.4% accuracy.

Adaptive mechanisms improved accuracy by 4.3%.

System operates with low latency and sparsity.

Abstract

We present an experimental imaging flow cytometer using a 1 {\mu}s temporal resolution event-based CMOS camera, with data processed by adaptive feedforward and recurrent spiking neural networks. Our study classifies PMMA particles (12, 16, 20 {\mu}m) flowing at 0.7 m/s in a microfluidic channel. Processing of experimental data highlighted that spiking recurrent networks, including LSTM and GRU models, achieved 98.4% accuracy by leveraging temporal dependencies. Additionally, adaptation mechanisms in lightweight feedforward spiking networks improved accuracy by 4.3%. This work outlines a technological roadmap for neuromorphic-assisted biomedical applications, enhancing classification performance while maintaining low latency and sparsity.