Human Brain Project, Caltech

The aim of this project at the California Institute of Technology is to use emergent computational technology to create models of brain development. The project team is a comprised of computational scientists, neuroscientists, and embryologists. Together they are using computers to analyse the results of optical and fMRI brain imaging. The long term goal is to provide an atlas of the complete anatomy and neural wiring of the brain, both in the adult and throughout childhood development. The project is split into three parts, as discussed below.

Algorithms project:
Understanding the structure and the development of the brain requires information through a range of scales. Global anatomy is at the scale of centimetres, whilst neural connectivity is at the scale of nanometers. Neuron signalling occurs within microseconds, brain development takes months. No single imaging technique, however, can be used at all of these scales. Optical imaging, for example, is capable of high spatial and temporal resolution but is unable to view deep inside of tissues. Conversely, MR imaging can non-invasively view deep inside living tissue but has much poorer resolution. Therefore computer melding techniques must be used to combine the data from different imaging modalities into a single image. Developing the software algorithms for this melding is the job of the Algorithms project.

As well as data-melding software, this team have also developed tools for tissue discrimination as well as hardware and software for semi-immersive visualisation of tissues.

Connectivity project:
This team uses the data from multi-modal imaging to analyse the anatomy of brain structures and their patterning of neuronal connectivity. The patterns and growth of axons and cell bodies are traced through the use of biochemical markers.

The key to combining images from the different techniques is biochemical markers which are visible to both optical and MR imagers. Novel markers (also known as contrast agents) have been developed. They are usually made up of two or more linked parts, an MRI active part and an optically active (fluorescent) part.

These markers can then be applied to neural tissue, either in vivo or in vitro. The markers will bind to the membranes of specific nerve cells. The growth of these cells and their axons is traced using various forms of light and MRI microscopy. The project has access to an 11.7 Tesla MR imager.

Atlas project:
The Atlas project disseminates the results of the above two projects to produce an in vivo atlas of the developing brain. Atlases are being produced not just of the human brain but also other species such as frogs, rodents and primates. These atlases are prepared for distribution via the web and on CD Rom.

Links:
The Human Brain Project at Caltech: www.gg.caltech.edu/hbp