VCSEL Based Time-Delayed Spiking Liquid State Machine for Event-Based Flow Cytometry Image Classification

Imaging flow cytometry (IFC) is a cellular analysis technique that combines the features of flow cytometry and confocal microscopy to provide rich morphological information for high-speed particles. Event-based cameras can address limitations of conventional CMOS detectors in IFCs, such as motion blur and high data volume. On the other hand, data generated by these IFCs are still processed through conventional deep neural networks that fail to exploit their asynchronous, spiking nature. In this work, we utilize an electro-optic processing paradigm based on a power efficient vertical surface emitting laser (VCSEL) excitable system [2], targeting the classification of experimental IFC data; consisting of high-flow particles of different diameters (fig. 1a) [1]. VCSEL powered spiking time-delayed liquid state machine (TD-LSM) processing offered an experimental classification accuracy of 96.4%, offering massive compression by a factor of 18, compared to standalone digital processing