Introduction

Magnetic resonance imaging (MRI) has been playing a vital role to study brain anatomy and functions and their abnormalities. Its advantage over other imaging modalities is obvious. It is non-invasive and can create various contrasts to study different types of brain tissues and functions such as white matter - gray matter differentiation, blood-brain-barrier leakage, angiograph, hemodynamics, to name a few. However, MRI has many limitations too. For example, MRI cannot clearly distinguish such important brain structures as cortical layers and thalamus nuclei. It is often insensitive to cellular abnormalities or degeneration, which can be readily detected by histology. Many of the limitations stem from two key factors in MRI; spatial resolution and contrast.

Imaging (spatial) resolution of MRI is a function of imaging time and signal-to-noise. The higher the resolution, the longer the scanning time and the lower the signal-to-noise. The signal-to-noise is in turn determined by hardware (most notably, magnetic field strength) and the type of imaging technique employed. In in vivo studies, imaging time is often a limiting factor. If we have infinite scanning time and the strongest magnet, we can achieve very high-resolution imaging. How high can it be? MRI is a technique to locate water molecules. During the measurement of typically 10 - 100 ms, water molecules inside the brain at 37 degree move 1 - 10 micron. In other words, MRI cannot locate the water molecules less than 1 - 10 micron, which is the physical limit of imaging resolution by MRI. In practical situations, the imaging time and/or signal-to-noise ratio usually hit the limit before the imaging resolution reaches the physical limit. That means, there are still a lot of rooms for technical development to enhance imaging resolutions using better hardware and clever imaging techniques.

The second issue is contrast. Conventional MRI such as T1 and T2-weighted images can do a great job to differentiate the gray and white matter. However, they often provide poor contrast to identify substructures within the gray or white matter; they look rather homogeneous. If anatomical units of interest look homogeneous, we cannot study structures within the unit no matter how high the resolution is. MRI contrasts are based on physical and chemical properties of water molecules. Depending on imaging techniques, the contrast is generated based on different properties of water molecules, such as rotational motion, translational motion, flow, and proton exchange process. Similar to computer, MRI needs software and depending on the software, it can do different tasks. With clever software, it can generate new contrasts to provide richer anatomical and functional information.

The goal of our laboratory is to develop new MR technologies to improve the resolution and contrast of MRI and apply them to observe brain anatomy to answer various types of biological questions. Currently we have three major research targets;
1: Characterization of mouse brain development Brain is by far the most complicated organ inside the body and its development is still beyond our understanding. However, recent advent of mammalian gene engineering, especially using mouse, now allows us to investigate molecular mechanisms of the complicated yet highly organized sequence of the development. In order to study the developing neuroanatomy, histology is almost exclusively the method of choice. It has high spatial resolution and the contrast technique is almost limitless. Histology can detect cells with specific chemical characters, locate genes and proteins, and investigate the sub-cellular structures. However, one thing it cannot do, or not practical in terms of labor and time, is macroscopic characterization of the neuroanatomy. For example, how the 3D shapes of ventricle and neuroepithelimum change during the development? Which gray matter structure is most affected by a gene alteration in terms of volume loss? MRI is a promising tool to macroscopically characterize the brain anatomy. In this project, we are pushing the envelope of image resolution and contrast technique to identify developing neuronal structures. To learn more, click here.
2: Human white matter anatomy and development This project is based on a new MRI technique, called diffusion tensor imaging (DTI). In the DTI technique, the directionality (anisotropy) of water diffusion in the brain is measured. This anisotropy is due to the orientation of bundles of axons and/or their myelin sheaths. Namely, brain water preferentially diffuses parallel to fiber bundles rather than perpendicular to them. Thus, the measurement of diffusion orientation carries interesting information about the brain anatomy that has not been accessible by conventional imaging techniques. In other words, DTI can provide new and powerful contrast to study brain anatomy, which cannot be obtained by other conventional MRI.
The overall aim of this study is to demonstrate the feasibility of generating computerized human brain atlas based on the DTI. Knowledge of neuronal connections by the white matter tracts is of critical importance for the understanding of normal brain functions and brain abnormalities. However, to date most approaches have relied on invasive in vivo techniques and, necessarily, human data have been severely limited. In this project, we have two long-term goals. These are elucidation of white matter architecture and studies of brain development.
Study of white matter architecture: The white matter consists of axonal tracts with convoluted trajectories and it has been difficult to delineate them using non-invasive techniques. Using the DTI, we expect that we can study the detailed anatomy of the human white matter and its abnormalities. There are two subaims in this project. First, human white matter atlas will be created using the DTI. Second, a DTI database for normal human brains will be created. The atlas and database will be accessible through this website, which will facilitate the development of various data analysis and visualization tools and contribute the understanding of the white matter anatomy.
Study of human brain development: Human fetal and embryonic brains have very poor MR contrasts and it has been difficult to study the development of various early structures using MRI. However, the DTI can provide rich anatomical information based on structural orientations. This makes the DTI a suitable tool to study the brain development. In this project, we will create a DTI database of developing human brains, which will be accessible through this website.
To learn more about this project, click here .
3: Development of diffusion tensor imaging technique and technology dissemination.
This is a technology-oriented project to develop new techniques for all fronts of diffusion tensor imaging; data acquisition, processing and analysis. For the data acquisition, we are working on hard and software development for DTI, which is faster, with higher signal-to-noise ratio, with less image distortion, and less motion sensitive. For the data processing, development of three-dimensional tract reconstruction and visualization is our important project. We are also working on C++ based software with user-friendly interface, so that the techniques become more available. For the data analysis, quantification of the fiber architecture, revealed by the DTI is our main target. These projects are dedicated to 1.5 and 3.0 T clinical scanners.
To learn more about this project, click here .