99 of 102 people found the following review helpful
on December 8, 2010
The label "networks" in the title of this impressive book may fail to fully capture the incredible richness of intricate (multi-scale) brain structure depicted here. Perhaps an added adjective like "genuine" might serve to contrast this work with the many "toy" neural networks illustrated in other publications. To get some idea of the complexity of the genuine brain networks discussed here, picture your living room fully packed top to bottom with centimeter scale worms representing (scaled up) axons linking 100 billion cell bodies. Your worm visitors occupy multiple intricate paths; many are short and form local worm societies (modules), but some cross the entire room and allow remote worm modules to interact non-locally.
Sporns asks what network science might tell us about the brain. He begins with a non mathematical overview of graph theory, "graphs" being mathematicians' abstract label for "networks." Sporns considers both structural (fixed wiring) and functional (dynamic) interactions between brain network nodes and modules. "Modules" are defined here as communities of nodes with large numbers of internal interconnections that may, in some cases, be viewed as "super nodes" or nodes defined at larger scales. To adopt the metaphor of human social networks, neurons are analogous to persons and the modules at various scales are analogous to neighborhoods, cities, and nations.
Like social systems, brain networks exhibit a striking (nested) hierarchical modularity, essentially small networks within larger networks within still larger networks, much like nested Russian dolls. This multi-scale structure may account for much of the brain's complex behavior. I quote the famous neuroscientist Vernon Mountcastle with Sporns' provocative suggestion added in brackets, "the dynamic interaction between brain subsystems [organized in modular hierarchies] lies at the very essence of brain function." Sporns emphasizes this point by pointing out that descriptions of the brain at large scales should not be regarded as poorly resolved approximations of an underlying microscopic order; rather different scales offer parallel and complementary views of brain organization. Failure to appreciate this critical issue and focus only on a single favored level of organization may be labeled "scale chauvinism" (my words).
One important idea emerging from graph theory is that of "small world" networks, illustrated in social networks by strangers (perhaps living on opposite sides of the world) linked by a few acquaintances. The high density of short-range brain connections coupled with a small admixture of long-range connections favors small world behavior. Small worlds also promote high complexity; they appear to be quite abundant in brain structural networks, across systems, scales, and species. Network disruptions, perhaps due to lesions of network hubs, are believed to be associated with mental disturbances or other diseases.
A later chapter focuses on the neural complexity issue addressed in several earlier chapters. While there is no agreed upon rigorous measure of neural (or any other system) "complexity," many complex systems have certain common features, including the hierarchical modularity evident in brain tissue. Sporns argues that system complexity is high when order and disorder coexist. For example, the molecules in a gas exhibit (random) disorder, whereas the molecules in a crystal are ordered, but neither system qualifies as a complex system. Rather, organizational mixtures of order and disorder are hallmarks of complexity. Another common feature of complex systems is that segregation and integration of structure and dynamic activity coexist. Different parts of the brain do different things; yet they work together to produce a uniform behavior and consciousnes, a condition greatly facilitated by small world networks.
This book should have broad appeal among many neuroscientists working in disparate areas of brain science. The writing is clear with many useful figures (including beautiful color plates) and directed examples absent even a single equation. The latter feature will evidently broaden the book's appeal, although some may wish for some mathematical support in an Appendix. In any case, one can confidently predict that Sporns' book will become an essential reference on many neuroscientists' bookshelves well into the future.
The material in this book overlaps several other books aimed at broad audiences. Earlier in his career, Sporns worked closely with Gerald Edelman and Giulio Tononi; their book A Universe Of Consciousness How Matter Becomes Imagination (1995) provides and nice introduction to Sporns' conceptual framework in "Networks of the Brain." My new book (2010) emphasizes the critical importance of nested hierarchy in brain tissue, and also speculates in the wider world of intra and extra cranial information and its possible fundamental role in the production of consciousness.
34 of 35 people found the following review helpful
on January 11, 2011
This book is an excellent coverage of brain networks, covering structural, functional, and effective connectivity and their respective dynamics. I am a neurobiologist but this book is presented in such a way that it builds up to complex subjects with topics and language that non-neuroscientists (for instance computer scientists and mathematicians) can understand. There are no equations in the book and yet the mathematical concepts are explained clearly so that they can be easily understood.
22 of 25 people found the following review helpful
on March 2, 2011
I'm not a neuroscientist but have read a fair amount on the subject. This book was great, it builds up an understanding of networks in general and their application to brain science. It summarizes a lot of the latest science and makes claims that are largely in line with other authors I've read (e.g. Tononi, Gerald Edelman, Joaquin Fuster). In other words, from the perspective of someone on the sidelines, Sporns backs up the central claims with a whole lot of evidence. I'm very glad I read it and I plan on re-reading it soon to make sure it sinks in. Whatever little networks make up my brain got a good workout
13 of 15 people found the following review helpful
on August 20, 2011
This appears to be a fascinating book, but the Kindle version does not include the color plates, which is very disappointing, given that you must pay more for the Kindle version than a hard copy. I returned my Kindle version, and will be buying the hard copy instead.
After buying a hard copy of the book, I discovered that the color plates are indeed to be found in the Kindle version, at the very end of the book. It is unfortunate that Amazon does not include information about color illustrations in their online descriptions--the sample chapter that you can download for free does not always give any indication of whether color plates are included (according to the Amazon representative I spoke to, it's up to each publisher to decide whether or not to include color plates in the Kindle version).
In the end, I returned the hard copy and repurchased the Kindle version, because I mostly use Kindle on my Mac, which allows me to easily copy text and paste it into another document, saving me many hours of typing and editing by hand.
7 of 7 people found the following review helpful
on June 25, 2011
I work in neuroimaging. I've read this book cover to cover and I'm very impressed with the clarity of the writing.
Sporns tells you where he's going and summarizes each section so you are sure to get the main points. Each section leads into the next section.
The only chapter that did not provide enough background for me was Chapter 12 on Dynamics....clearly a difficult topic.
Thankyou Dr Sporns! You've opened up a whole new world to me.
3 of 3 people found the following review helpful
on December 5, 2014
This book primarily describes how brain connectivity can be analyzed and understood using graph theory and some related areas of study. The book begins by arguing for a graph theory approach to neuroscience. This is followed by some basics in graph theory, types of brain networks and neuroanatomy. The book then spends some times considering brain efficiency from an evolutionary standpoint, spontaneous neural activity (absence of stimulus), states of behavior/cognition, disease, brain development, dynamic networks, and complexity. Finally, the books closes by suggesting that the brain is not isolated and requires body and environmental feedback to be properly understood.
There is a massive amount of information contained in this book and Sporns has done a fantastic job conveying this information in a clear and organized manner. There are hundreds of cited examples, studies, and statistics to back up all the claims made by the author. However, this book is extremely dense and packed with information. Having finished the entire thing, I cannot remember nearly half of what I've read in great detail. I think I would treat this book as partially a reference guide - you can think of each chapter as a review article.
The reason I have removed one star is because the book begins by outlining its goal as a non-mathematical approach to brain networks for those outside the field. However, I think some parts of the book would still be quite difficult for someone with no knowledge of computational neuroscience. I research primarily in the molecular neuroscience field but I have familiarity with concepts in bioinformatics, graph theory, statistical neuroscience, information theory, and attractor dynamics. With this basic understanding I was often able to get an intuitive understanding of the ideas without really understanding them in depth. But there were many cases where I wondered if someone with no computational background would really understand certain topics. At times Sporns skips over topics entirely (for example, attractor networks are not really defined) and at other times he devotes quite a few pages to some of the very elementary concepts of complexity theory (that a complex system is part random and part modular). So, it is not always entirely clear who the audience really is. I agree that with the amount of information to be covered it would not be feasible to give a background to everything. But at times I found myself wishing for a more complete explanation in some areas and less in others.
Overall, this is a minor point. The book is well-written, thorough, dense, and fascinating. Instead of getting bogged down in the details, think of it as a very fast ride through the world of brain networks that does not give you the time to step out to explore everything in a lot of detail. I can see myself returning to chapters of this book in the future to learn something new.
2 of 2 people found the following review helpful
on June 12, 2013
This book is a tour de force, an ambitious and thorough review of the field of network neuroscience that everyone interested in cognitive neuroscience or related fields should read. In short, Sporns shows how the tools, concepts and frameworks developed to analyze networks (say traffic control, communications, the world wide web, or whatever) can be used to analyze brain structure and function, from cellular levels all the way up to complex cognitive processing. Although it would seem at first intuitive (the brain being composed of a vast array of interconnected nodes, a network), the field is quite young, and most literature reviewed by Sporns is very recent (the last 20 years or so). This book will be seriously discussed by cognitive scientists for years to come.
In the first chapters Sporns defines network and information theory tools and concepts, such as "nodes", "edges", "clustering", "path length", etc. in non-technical terms (that is, no equations, although fully understanding everything in the book could require more than one read), and explained how they are used to analyze complex networks. The rest of the book is devoted to showing how these can be applied to brain networks, and lead to novel predictions and adequate modeling of brain structure, brain evolution and development, brain dynamics, cognition, brain functional disorders and behavior, to name a few. Just as an example, "small-world network", one with high clustering and short-path length turns out to be a property of complex, modular interconnected networks, and it is indeed also a characteristic of brain networks, that studies have correlated to efficient connectivity and cognitive processing.
Sporns does leave me with a thirst for some more speculative ideas. He does state that "perhaps network thinking will eventually allow us to move beyond neural reductionism and cognitive functionalism and formulate a theoretical framework for cognition that is firmly grounded in the biology of the brain" (page 206). But I would argue that the whole book is a big empirical argument against strong functionalism (mind is identical to the functional properties of abstract information processing, independent of any hardware, thus "multiple realizable"). A recurrent theme is how structure and function come hand in hand and how one determines the characteristics of the other, and not in a trivial way. These ideas lead to a very reasonable and moderate interpretation of embodied cognition (Chapter 14).
Network theory seems to lead to a fusion of functionalist and identity theory views of mind, where each dynamic state of a complex network depends on both a specific functional role and on a specific neural implementation, and cannot be solely reduced to either description. This more "philosophically" oriented interpretation could lead to a resolution of many modern debates in philosophy of mind (consciousness and qualia, for example; although Sporns does comment on the "dynamic core" hypothesis of consciousness, a network-derived theory of consciousness proposed by himself, Edelman and Tononi, involving integration and complexity measures of network dynamics as essential for conscious processing). Whether this specific proposal is on the right level of description to have explanatory value for the problem of consciousness is open to debate. Just because consciousness is complex and integrated does not mean its neural correlates must be complex and integrated. They could very well be, but it's just like saying "consciousness is mysterious, quantum physics are mysterious, and so they must be related". But I digress. As a fully empirical approach, network neuroscience will surely continue to grow and I predict soon will be successful in explaining some of the most complex issues in cognitive science.
4 of 5 people found the following review helpful
on August 5, 2011
I read "Networks of the Brain" from cover-to-cover, and looked forward to reading more of it after each new chapter. Olaf Sporns has crafted a beautiful little book. The writing is some of the best I have seen. His style is clear and elegant, and he makes it all seem effortless. It is so very refreshing to see.
on March 17, 2014
What really makes this book stand out for me is the caliber and detail of its extensive illustrations. Without them, the associated text would have been problematic to grasp readily or adequately for a non-specialist like me. Quite often here, individual illustrations have been partitioned into complementary views or logical sequences that greatly aid in solidifying the respective concepts or phenomena. While crediting the sources of particular illustrations, Sporns notes individually which of those figures have been redrawn. The resultant consistency of style considerably enhances reader discernment of their content. In all, this selection, composition, and editing of figures are exceptional as well as unique in my reading experience.
Given his extensive background in the subject or related research, the author has selected, integrated, and interpreted the significant findings of a broad range of ostensively esoteric research publications, and rendered them in a quite readable form. The fact that most of his references are journal articles, moreover, indicates the recency or provisional nature of much of the reported brain network characterizations and implications. Although the research contributions of Sporns himself are evident in the references cited, he nonetheless offers a balanced view that explores contending theses, interpretations, or uncertainties. I found his focus on networks, and their heterogeneous neural constructs and flexible dynamics, to be a vital and vivid unifying theme - from which his overall message issued quite coherently. Furthermore, his unified synthesis of network topology and dynamics constitutes the pivotal factor in a well-established framework for continuing research in brain science. Quite significantly, empirically established network (re)organizations and constructs are shown to be amenable to a significant range of revealing quantitative analysis methods and corroborative simulations.
Certain topics were of especial interest to me. In Chapter 3, three types of brain connectivity are introduced: structural or anatomical, functional or interactional, and effective or causal. These forms and their roles/relationships/patterns are then deployed or elaborated throughout the rest of the book. Also, the nature and contributions of spontaneous neural activity, as ongoing patterns of strictly internal brain operations, are an intriguing phenomenon developed in Chapter 8. Such activity serves as an active standby basis for mediating incident external stimuli. They resultantly prompt operative network reconfiguration and functionality - so as to produce suitable responses. Thereafter, the rather novel notions of metastability and coordination dynamics are presented in Chapter 12. Metastability inheres in a quasi-repetitive state transition trajectory in conformity with a prevailing yet tenuous attractor manifold. This tentative trajectory reconciles the opposing tendencies of brain element segregation and integration, while maintaining readiness for rapid task-induced configuration as entailed in the onset of coordination dynamics. When invoked, such coordination effects the coupling of certain modules and thereby enacts transient process integration. This in turn results in appropriate cognitive patterns to yield indicated functional performance.
Fundamental yet discreet reservations concerning several long-standing viewpoints or speculations still prominent in cognitive science literature appear at relevant points in the book. Sporns' observations refer to concepts/initiatives that are simply incompatible with extant brain science, especially with respect to the brain network paradigm he articulates. (Use the term paradigm here, rather than framework, would seem to be warranted by the scope, coherence, conclusiveness, and maturity of the combined empirical and analytical support that is consolidated in the overarching brain network theory he presents.) The following points are of particular import:
· "If cognition is largely symbolic in nature, then its neural substrate is little more than inconsequential detail revealing nothing about the essence of the mind." (p.179)
· "neuroreductionism...implies that cells and molecules can explain all there is to know about mind and cognition...yet it cannot explain their emergent and collective properties." (p. 180)
· "Structural and functional modules identified by network analysis have little in common with the putative cognitive modules proposed by Fodor...Such computational modules...are restricted to its (innate) database,...and...are cognitively impenetrable by other extramodular processes." (p. 195)
· "connectionist approaches...do not address how network growth and plasticity shape the emergence of neural dynamics...(their) models bear little resemblance to the complex architecture of the brain." (p. 245)
· "'The ant, viewed as a behaving system, is quite simple'"(p. 308)...(yet) "ants are complex organisms with complex nervous systems...(which implausibly) `may be irrelevant to the ant's behavior in relation to the outer environment.'" (p. 345)
The book's exceptional merit notwithstanding, I believe that a few matters might be treated more clearly. First, some readers would be well served by the judicious use of tables - to consolidate or summarize vital points or distinctions among the diversity of comparable topics or techniques addressed. For example, tables that outline the nature, scope, resolution, and practical utility of various imaging techniques and simulation/analysis methods would be useful checkpoints for non-specialists. Second, a taxonomy of network forms would be helpful, along with a tabular summary of their salient attributes and their typical roles. Then, it would be of benefit to have: a table that characterizes typical network elements (e.g., map, module, motif); figures that depict generic constructs (e.g., recursion, reentrancy, motifs); and incisive content in all such cases regarding each element's respective applicabilities and typical locales within the central nervous system. These are surely not major concerns, but merely prospects for potential improvement to an already excellent exposition.
In sum, this book serves motivated non-specialists as an authoritative and unified technical literature critique and synopsis - of an otherwise virtually impenetrably complex of research threads and findings. Moreover, it might well be valuable for the orientation of beginning students in cognitive science. Be assured that this book is not science popularization or hype as sometimes typical of mere science writers. Rather, it is a sober and penetrating illumination of brain networks in an expansive and unified sense. In general, the book's organization and its balance of coverage are very good; moreover, the endnotes are ample and most helpful. In all, the book's content is substantive and coherent, encompassing yet cohesive, impressively grounded, and ultimately quite persuasive. Since the author's native language is apparently not English, the book's caliber of writing is especially remarkable. I rate this book at 5-Plus Stars.
2 of 3 people found the following review helpful
on February 8, 2014
This is an excellent book. I've had it for a while and finally got around to reading it. It took me a long time because it is rich with detail and is a captivating read. This is not a book for the novice with a embryonic curiousity in neural networks. There are other books that will do a much better job at introducing you to this field. But if you have been in it for a while, you won't be dissapointed. I have PhD training in computational neuroscience and even I found this book rather densly written. I had to read it slowly. So if you don't have the background, you will read references to non-linear dynamical systems, or information theory, that will be lost on you as you read past. The book assumes some knowledge so name drops without explanation and expects you to understand the reference and keep up. It does not go too far in its assumptions, however. It is not a text book so does not force you to become Ramanujan as you read it. It is a nice balance between a novice book and an advanced text. A must have.