News & Analysis
/
Article

A better platform to image neuronal activity in brain development models

JUN 18, 2026
Combining microscopy, scanning, and deep learning enables more precise imaging of functional dynamics in neural networks of human cortical organoids.
Hannah Daniel Headshot.jpg
Media Relations Specialist
A better platform to image neuronal activity in brain development models internal name

A better platform to image neuronal activity in brain development models lead image

In lieu of studying a full brain, stem cell-derived human cortical organoids (hCOs) have emerged as simpler models for studying brain development. Because the tissues are 3D, normal flat imaging cannot capture the complex dynamics of hCO cells, so volumetric calcium imaging is used to monitor neuronal activities.

Traditional light-field microscopy (LFM) can image large volumes quickly; however, retrieving spatiotemporal calcium activities of hCOs for long-term analysis is challenging since the technique suffers from out-of-focus background noise in thick volumetric specimens. Min et al. combined multiple approaches to better monitor calcium.

“Our approach preserves the single-shot volumetric advantage of LFM while using annular scanning to effectively reduce background noise via the illumination and collection process,” said author Moonseok Kim. “Therefore, it is especially useful for sustained, low-phototoxic functional imaging of organoid-scale specimens where scan-free volumetric acquisition is important.”

The researchers’ platform combines three techniques: LFM to capture the 3D calcium dynamics; annular scanning (AS) to increase the accuracy of LFM and reduce out-of-focus fluorescence; and finally, deep learning algorithms to remove image noise.

“Together, AS-LFM, the optical and computational imaging platform allows low-phototoxic, scan-free volumetric calcium imaging with several-fold enhancement of signal-to-noise ratio in live human cortical organoids,” said Kim.

The researchers see their platform being used to minimize optical aberrations and extend imaging depth. They plan to improve the processing time of their deep learning algorithms.

“Biologically, we aim to apply this platform to larger organoid datasets, disease-model organoids, and drug response studies, where long-term 3D functional readouts could provide more informative measures of human neural circuit development and dysfunction,” said Kim.

Source: “Noninvasive rapid observation of volumetric calcium transients in human cortical organoids using light field microscopy with annular scanning,” by Cheol Hong Min, Seungwon Jeong, Jeong-Sun Choi, Sangjun Park, Hyun-jong Jang, Kyung-Ok Cho, and Moonseok Kim, APL Photonics (2026). The article can be accessed at https://doi.org/10.1063/5.0332998 .

More Science
/
Article
Simulations show no drop on crossed-beam energy transfer compared to linear-polarized light but significantly reduced beam-to-beam power variations, pointing the way to more powerful lasers for fusion ignition.
/
Article
Two equations and one free parameter provide a surprisingly good mapping from laser to hohlraum radiation temperature pulses.
/
Article
Oscillatory retinal neuron networks don’t require external voltage sources and show comparable performance to cutting-edge GPU-based convolutional neural networks, for energy costs thousands of times less than current approaches.
/
Article
The rotational speed of the cutterhead used to excavate a tunnel plays a key role in the passageway’s stability during the excavation process.