A Derivative-Based Framework for Real-Time Signal Processing and Event Detection in Impedance Flow Cytometry.

Impedance flow cytometry (IFC) enables label-free, real-time characterization of cells and particles, but its performance depends critically on accurate event detection and feature extraction under varying noise and acquisition conditions. Conventional pipelines typically rely on multi-stage thresholding, wavelet transforms, template-based correlation methods, or neural-network models. These approaches generally require additional preprocessing steps and involve multiple parameters or hyperparameter tuning. In this work, we present a simple derivative-based signal processing framework that enables baseline-drift suppression, event detection, and feature extraction within a single computational step. The derivative approach improved precision and recall by approximately 20% and reduced the false discovery rate by 15-25% compared with simple thresholding, while requiring only 22-55% of the processing time across all the test conditions. The algorithm operates in linear time with minimal memory overhead and does not rely on template matching or trained parameters, making it well-suited for real-time or embedded, resource-constrained IFC platforms. We further demonstrate that derivative-extracted features enable accurate real-time classification of microparticles, achieving >98% accuracy while maintaining a processing speed that is approximately two orders of magnitude faster than the data-acquisition rate.