Scanning Brain Networks
Refi nements of these models by using the high-resolution data should reveal fi ner 3D aspects of the fl y brain. Human brain circuits . Figures 4 and 5 show the structures in the human frontal cortex tissues. T e 3D images of a block sample with dimensions of 1.4 mm×1.4 mm×8.3 mm are so complicated ( Figure 4 ) that they cannot be comprehended at a glance, although pyramidal neurons arranged in a soma layer, called the internal pyramidal layer, can be seen in the image. A smaller sample with dimensions of 0.30 mm×0.35 mm× 2.4 mm was also visualized with tomographic microscopy and subjected to model building. Skeletonized models of neurons ( Figure 5a ) were built by tracing the neuronal processes and assembling them into neurons. Although it is still diffi cult to comprehend the entire models, any of the neurons composing them can be extracted ( Figure 5b ). Because the models are described in terms of 3D Cartesian coordinates, the distances between the neuronal processes or somas can be directly calculated from the coordinates. T is enabled us to determine individual neuronal circuits by analyzing the positional relationships of the neurons [ 1 ]. T e pair of neurons shown in Figure 5b connect their inputs and outputs to each other to form a feedback loop ( Figure 5c ). In electronics, a similar feedback circuit composed of transistors is known as an astable multivibrator ( Figure 5d ), which generates a string of pulses. T e refractory period of neurons prevents repetitive action potentials and hence poses a limit on the fi ring interval. However, a delay through a loop composed of a number of neurons allows recovery from the refractory period of a few milliseconds, resulting in oscillations of the loop circuit. Because such feedback loops are formed if a number of neurons are connected to each other, the loop circuit should be one of the canonical structures of human brain circuits.
Figure 4 : Three-dimensional structure of human frontal cortex tissue. X-ray linear attenuation coeffi cients were rendered from 9 cm -1 (white) to 112 cm -1 (black). Arrowheads indicate the internal pyramidal layer. Scale bar = 100 µm.
T e network constituents were classifi ed into groups on the basis of their 3D structures [ 2 ]. T e classifi ed model ( Figure 3 ) allowed us to extract anatomical segments by specifying neuronal processes in a group-by-group manner. For example, the lobula plate ( Figure 3e ) located at the posterior end of the optic lobe can be distinguished from the other structures by its location and structures specifi c to the lobula plate. Neurons of the lobula plate were characterized by their fan-shaped ramifi cation. Such structures of the lobula plate can be extracted from the model by specifying corresponding structural groups. Figures 3 b– 3 f show the organization of the optic lobe network, which is responsible for visual information processing. Neuronal processes of the medulla and second optic chiasma, which are proximal to the compound eye, exhibit periodical structures corresponding to repeated units of photoreceptors. On the opposite side of the optic lobe, neuronal processes are assembled into several tracts pointed toward central brain regions. T ese structures represent information paths from the inputs on the eye to the integrative process in the central brain.
2015 September •
www.microscopy-today.com
Discussion T ree-dimensional structures of neuronal networks of human and fruits fl y brains were visualized with X-ray tomographic microscopy [ 1 , 2 ]. T e obtained structures were analyzed by building skeletonized models of neurons. T is quantitative description of 3D networks in Cartesian coordinate space enabled examination of the functional mechanisms of the obtained brain networks. T e functional architecture of the visual cortexes of cat and monkey brains has been incorporated into artifi cial neural networks [ 4 ]. T is neural network, called a convolutional neural network, has been applied to image recognition tasks and has outperformed other computing methods designed without reference to the visual cortex architecture [ 4 ]. T e present study illustrated some of the circuits of the human frontal cortex, which is responsible for higher brain functions. T ese circuits should provide a basis for cognitive computing to deal with complicated problems that have not been able to be handled without human intervention. T e psychological individuality of human beings is ascribable to the individuality of their brains. T erefore, psychological characters reside in the neuronal circuits of the brain. Diff erences between neuronal circuits should be observable not only in healthy individuals, but also between patients with brain disorders and healthy controls. Such diff erences can be clues to improving the treatments of brain disorders. Because our knowledge of the circuits in the human brain is still limited, drugs for treating
15
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44 |
Page 45 |
Page 46 |
Page 47 |
Page 48 |
Page 49 |
Page 50 |
Page 51 |
Page 52 |
Page 53 |
Page 54 |
Page 55 |
Page 56 |
Page 57 |
Page 58 |
Page 59 |
Page 60 |
Page 61 |
Page 62 |
Page 63 |
Page 64 |
Page 65 |
Page 66 |
Page 67 |
Page 68