MIT scientists scanned a larval zebrafish brain 200 times per second; ultraviolet light triggered electrical activity that spread across its brain

Imagine trying to understand how a city functions by watching one traffic light at a time. That’s essentially how neuroscientists have long studied the brain, a handful of neurons at a time. A team at MIT has now built a new high-speed microscope that images activity across nearly the entire brain of a larval zebrafish at once, scanning it every five milliseconds, roughly 200 times a second, and tracking how electrical activity spreads after a pulse of ultraviolet light.

Why neuroscience has been stuck watching reruns, not live TV

Most brain-imaging techniques rely on calcium imaging, which tracks calcium ions entering neurons after they fire. The problem: calcium signals unfold over seconds or minutes, far too slowly to capture how neurons actually compute. “It’s very slow: it takes seconds or minutes to get information with calcium imaging,” study co-author Jie Zhang told MIT News. Because individual neural computations happen on a millisecond timescale, that speed simply wasn’t enough to see them.

Voltage imaging, which tracks electrical activity directly rather than its calcium aftermath, has existed for some time, but earlier systems could only image small clusters of neurons in one region at once. One early landmark, a 2013 light-sheet imaging study from the Howard Hughes Medical Institute, captured roughly 80% of neurons in a zebrafish brain, but only via calcium signals, at less than one scan per second, as per Ahrens et al., in their 2013 paper titled “Whole-brain functional imaging at cellular resolution using light-sheet microscopy,” published in the journal Nature Methods. The MIT team’s new system improved on that scanning speed by more than 200-fold.

How it works

The team, led by Ed Boyden and reported in a separate 2026 paper titled “Voltage imaging of neurons distributed across entire brains of larval zebrafish,” published in Nature Methods, improved a type of microscope known as a light sheet microscope. These use a thin sheet of laser light to scan through the sample, creating 3D images. Faster cameras and a technique called remote refocusing let the system scan much more quickly than earlier designs.

The neurons within the fish were engineered for pan-neuronal expression of a genetically encoded voltage indicator called Positron2-Kv, which fluoresces when a neuron fires. Such indicators have long been considered a “holy grail” of neuroscience, per a 2019 review titled “Genetic voltage indicators,” published in the journal BMC Biology, for many years these have been considered the "holy grail" of neuroscience.

Only around 25% of neurons in the fish responded, providing useful data. This is an early version, so they are honest that this is not perfect. However, this was enough to show activity firing in the optic tectum (the visual processing centre of the brain) shortly after the UV exposure. This then spread across the brain, causing separate spontaneous activities in the cerebellum and hindbrain.

Toward mapping whole brains, not just zebrafish

You won’t be getting a photo of your neurons firing anytime soon. Zebrafish have small brains that are transparent and easy to manipulate using genetics. Human and mouse brains are not so easy to do this with. Whilst they are working on expanding their research to mice, the biggest benefit of this research is that we can finally see how neurons work together in a network. This reminds us that neuroscience is still catching up with the basics in this age of brain optimization.