by New AI Technology Enables Tracking of Neurons in Moving Animals
Scientists from EPFL and Harvard have developed an innovative artificial intelligence (AI) method to track neurons in animals that are moving and deforming. This breakthrough technology, based on a convolutional neural network (CNN), allows for the decoding of circuit activity in the brain, even while the organism is in motion. The study, led by Sahand Jamal Rahi from EPFL’s School of Basic Sciences, has been published in Nature Methods.
Traditionally, This becomes even more complex when the brain itself moves and changes shape within the organism’s flexible body. The lack of appropriate tools to address this problem has hindered progress in this field.
The new AI method utilizes a CNN, which has been trained to recognize and understand patterns in images. The CNN utilizes a process called convolution, where it analyzes small parts of the image at a time, such as edges, colors, or shapes, and combines this information to identify objects or patterns.
One of the main difficulties in tracking neurons during a movie of an animal’s brain is the need for manual labeling of many images. This is due to the fact that the animal appears differently over time due to various body deformations. Generating a sufficient number of annotations to train a CNN manually can be a daunting task, given the diversity of the animal’s postures.
To overcome this challenge, the scientists developed an enhanced CNN with targeted augmentation. This innovative technique automatically synthesizes reliable annotations for reference using a limited set of manual annotations. The CNN learns the internal deformations of the brain and uses this information to create annotations for new postures. This significantly reduces the need for manual annotation and double-checking.
The versatility of the new method allows for the identification of neurons represented as individual points or as 3D volumes in images. The researchers tested the technology on the roundworm Caenorhabditis elegans, a popular model organism in neuroscience with 302 neurons.
Using the enhanced CNN, the scientists were able to measure the activity in some of the worm’s interneurons, which are responsible for bridging signals between neurons. They discovered that these interneurons exhibit complex behaviors, changing their response patterns when exposed to different stimuli, such as periodic bursts of odors.
To make the CNN accessible to other researchers, the team has provided a user-friendly graphical user interface that integrates targeted augmentation, streamlining the entire process from manual annotation to final proofreading.
According to Sahand Jamal Rahi, the new method significantly reduces the manual effort required for neuron segmentation and tracking, increasing the analysis throughput threefold compared to full manual annotation. This breakthrough technology has the potential to accelerate research in brain imaging and deepen our understanding of neural circuits and behaviors.
Note:
1. Source: Coherent Market Insights, Public sources, Desk research
2. We have leveraged AI tools to mine information and compile it
