Brain-inspired Intelligent System
Knife-Edge Scanning Microscope Brain Atlas went live.
- Computational Neuroscience and Neuroanatomy
- Sensorimotor Learning, Temporal aspects of brain function
- Neural Networks, Neuroevolution, Machine Learning
- Embedded System Software, Embedded Application Platform
Broad Research Area: AI / MACHINE LEARNING / ROBOTICS / VISION, SCIENTIFIC/MEDICAL INFORMATICS
- Brain Networks Laboratory (at Texas A&M), September 2005-Present.
- Dr. Yoonsuck Choe (Computer Science and Engineering at Texas A&M)
Brain Image Acquisition and Analysis
The brain may be the final frontier of science. Scientists may have much more knowledge about the universe than their own brain. As brain science made great advances during the past few decades, we have come to understand the brain much more than any other time in human history. However, in spite of numerous significant findings, there still are many mysteries of the brain that have yet to be solved.
In order to understand the function of the remarkably complex networks of neuronal cells and vasculature, their elements and inter-connections should be investigated. However, the sheer structural complexity of the brain prevents us from studying brain function through examining its large-scale structure. For example, the human brain has approximately 10^11 neurons and their possible connections are estimated to be 10^14. Besides, the total length of the vasculature in the human brain is estimated as hundreds of miles and its surface area is more than 100 square feet.
It is a tremendous challenge to understand brain function because elements and interconnections of the brain networks should be fully investigated, given the sheer number of elements and their connections.
Recent advances in serial sectioning microscopy such as the Knife-Edge Scanning Microscopy (KESM), a high-throughput and high-resolution physical sectioning technique, have the potential to finally address this challenge. Nevertheless, there still are several obstacles remaining to be overcome.
- First, many of these serial sectioning microscopy methods are still experimental and are not fully automated.
- Second, even when the full raw data have been obtained, morphological reconstruction, visualization/editing, statistics gathering, connectivity inference, and network analysis remain tough problems due to the unprecedented amounts of data.
I designed a general data acquisition and analysis framework to overcome these challenges with a focus on data from the C57BL/6 mouse brain. Since there has been no such complete microstructure data from any mammalian species, the sheer amount of data can overwhelm researchers. To address the problems, I constructed a general software framework for automated data acquisition and computational analysis of the KESM data, and conducted two scientific case studies to discuss how the mouse brain microstructure from KESM can be utilized.
I expect the data, tools, and studies resulting from this research to greatly contribute to computational neuroanatomy and computational neuroscience.
Automatic Region Cropping and Artifacts Removal
KESM is unique in that illumination and tissue ablation are performed using a diamond knife. Therefore many of the physical forces applied to the knife (e.g., vibration, slip, and light refraction) manifest as image artifacts that must be removed in post-processing.
Fully automated framework to extract valid data from imaged sections are a crucial step before actually analyze internal structure of tissues. Tissue regions in a raw image should be automatically cropped. After that, fast and efficient artifact removal allows us to reconstruct the volumetric structure.
The first example is a image volume stained with India ink. This shows a unprocessed volume that have many artifacts that prevent us from properly analyzing internal structure (left). The right one shows the processed volume (note that this is as same region as above).
Here are blood vessels from other processed regions stained with India ink.
The second example is from a Golgi stained specimen. The volumes in the left column shows original raw images, and the right column shows processed volumes.
3D Reconstruction of Neuron Morphology and Microvasculature
To see the original image gallery, visit Multimedia Gallery page in Brain Network Lab website.
The left image shows Cerebella Purkinje cells from our whole brain Golgi data sets. Cortical pyramidal cells from our whole brain Golgi data set are shown in the right image.
Following images are tracing examples of fibrous structure. The images in the left column represents vasculature visualization. The lines in the right column shows the traced results.