Characterization of Event-Based Camera for High-Speed Recording of Cell Activity

Abstract

Event-based sensors use Address-Event Representation (AER) to detect changes in light brightness rather than capturing frame-by-frame images, allowing them to capture rapid events. This research assessed their suitability for cell activity and more specifically toward the end goal of voltage imaging, requiring 10% ΔF/F sensitivity and favorably better than 1 ms temporal resolution. Two different sets of tests were conducted to determine the maximum frequency and sensitivity of the sensor. The sensor demonstrated the ability to detect events over 1 kHz and irradiance in the nW/mm² range, needing brighter signals for faster detection. Imaging of beads and neuron cultures was conducted using a wide-field fluorescent microscopy setup with 2.4 μm resolution and a field-of-view of approximately 630 x 470 μm. Changes were detected via stage movement or light frequency modulation. Data were processed with noise filters and image processing techniques to retrieve signals from event data, confirming the camera's efficacy for voltage imaging. The study concluded that while the event-based camera shows promise, further refinement in noise filtering and image processing is necessary. These advancements could enhance temporal resolution to reveal fine details such as subthreshold potentials, aiding in drug toxicity testing, drug discovery, and neurodegenerative research.