Image Processing Reference
In-Depth Information
tion matrix is usually visualized using a color-coded matrix view. For more details
and examples see, e.g., the recent reviews [
12
,
30
].
In contrast to genomics, the field of connectomics is to a large extent based on
image data. Therefore, visualization of image data can directly support the analysis
of brain structures and their structural or functional connections.
In this chapter, we review the current state-of-the-art in visualization and image
processing techniques in the field of connectomics and describe some remain-
ing challenges. After presenting some biological background in Sect.
21.2
and an
overview of relevant imaging modalities in Sect.
21.3
, we review current techniques
to extract connectivity information from image data at macro-, meso- and microscale
in Sects.
21.4
-
21.6
. Section
21.7
focuses on integration of anatomical connectivity
data. The last section discusses visually supported analysis of brain networks.
21.2 Biological Background
Neural systems
. Functionally, neurons (or nerve cells) are the elementary signaling
units of the nervous system, including the brain. Each neuron is composed of a cell
body (soma), multiple dendritic branches and one axonal tree, which receive input
from and transfer output towards other neurons, respectively. This transfer is either
chemical (synapses) or electrical (gap junctions). Generally, during synaptic trans-
mission, vesicles containing neurotransmitter molecules are released from terminals
(boutons) on the axon of the presynaptic neuron, diffuse across the synaptic cleft,
and are bound by receptors on dendritic spines of the postsynaptic neuron, inducing
a voltage change, i.e., a signal.
These basic building blocks can mediate complex behavior, as potentially large
numbers of them are interconnected to form local and long-range neural microcir-
cuits. At the meso-level, local neuron populations, e.g.,
cortical minicolumns
, can
be identified that act as elementary processing units. At the macroscale, neurons in
the human cortex are arranged in a number of anatomically distinct areas, connected
by interregional pathways called
tracts
[
89
].
Model systems
. An important neuroscientific goal is to understand how the human
brain works. However, due to its complexity (with an estimated 10
11
neurons with
10
15
connections [
89
]), brain function at the circuit or cellular level is often studied
in other organisms that are more amenable in complexity and size.
Conserved genes and pathways between different species offer the potential
of elucidating the mechanisms that affect complex human traits based on similar
processes in other organisms. This problem is particularly tractable in the round-
worm
Caenorhabditis elegans
, whose brain with 302 neurons has been completely
mapped [
106
], or in insects. In these organisms brain structure and function can be
studied at the level of single identifiable neurons. Classical insect model organisms
that are well understood and allow easy genetic manipulations are the fruit fly
Drosophila melanogaster
and the honeybee. Drosophila, for example, has been
Search WWH ::
Custom Search