Event camera captures fast cortical blood flow

A project group including Seoul National University and Korea's KAIST research institute has applied an event camera to the imaging of cortical blood flow.

Discussed in PhotoniX, the results could point towards new kilohertz-rate vascular imaging capabilities and functional reconstruction in the living brain.

Event cameras, also called neuromorphic imagers, are designed to capture per-pixel changes in intensity, rather than full frames at fixed intervals as conventional cameras do. This approach is analogous to the way that the human eye perceives visual information, and event cameras are classed as bioinspired imaging technology.

But applying event cameras for functional brain imaging presents a very different challenge, commented the project, one in which the signals of interest are subtle changes in fluorescence tied to blood flow or neuronal activity. In event cameras each pixel independently reports only changes in brightness, a sparse stream of asynchronous events with sub-millisecond timing and a wide dynamic range.

These properties are attractive for functional imaging in principle, but they had not previously been validated in vivo for small-amplitude activity signals. Whether an event camera could reliably detect such small brightness changes, and do so in a way that is scientifically useful, has been an open question.

"Brightness changes arising from functional indicators are inherently smaller than those encountered in motion or blinking applications," said the project in its paper. "Illumination intensity is further constrained by photobleaching and phototoxicity, limiting attainable signal amplitude and signal-to-noise ratio."

The researchers tackled the problem through an event-based imaging framework combining quantitative optical characterization of event cameras, multi-modal in vivo imaging including cortical blood flow and neuronal calcium dynamics, and a novel self-supervised reconstruction algorithm termed Implicit Neural Factorization (INF) which converts sparse event streams into continuous activity signals.

Advantages of event cameras

An initial experiment characterized the sensor's response to functional signals rather than large object motion, to see how reliably the camera reported small, slowly varying changes in brightness and evaluate noise behavior and sensitivity relevant for biological imaging. This calibration provided a key link between the binary events recorded by the neuromorphic sensor and the underlying changes in intensity that neuroscientists interpret as blood-flow or calcium signals.

With this link in place, the event camera was applied to the monitoring of cortical vascular dynamics in mice. Comparing the event streams to conventional widefield measurements showed that the sensor could faithfully track fast changes in blood flow signals. Importantly, the vascular activity was captured at an effective rate of 1 kHz, highlighting that event-based acquisition can reach kilohertz temporal resolution while covering a large field-of-view.

The same framework was also applied to neuronal calcium dynamics, in cultured neurons and in vivo mouse cortex, with the INF algorithm turning binary asynchronous measurements into functional data. For this neuronal calcium activity, INF allowed a direct comparison with conventional sCMOS recordings while still benefiting from the data efficiency of event-based acquisition, commented the project, and results indicated that the combination of wide field of view and 1000 Hz sampling was a regime not readily accessible to conventional cameras.

"The goal of this work was to establish that event-based sensors can be effectively applied to in vivo functional biological imaging, a context in which sCMOS cameras are widely used," wrote the team. "Even in regimes where sCMOS systems may approach similar frame rates, event-based sensors provide practical advantages, including compactness and low power consumption, and are beneficial for portable functional imaging platforms."