What does fMRI tell us about brain homologies?
Friday, May 11, 1:00 – 3:00 pm
Organizer: Reza Rajimehr, McGovern Institute for Brain Research, Massachusetts Institute of Technology
Presenters: Martin Sereno, Department of Cognitive Science, UC San Diego; David Van Essen,Department of Anatomy and Neurobiology, Washington University School of Medicine; Hauke Kolster,Laboratorium voor Neurofysiologie en Psychofysiologie, Katholieke Universiteit Leuven Medical School; Jonathan Winawer, Psychology Department, Stanford University; Reza Rajimehr,McGovern Institute for Brain Research, Massachusetts Institute of Technology
Over the past 20 years, the functional magnetic resonance imaging (fMRI) has provided a great deal of knowledge about the functional organization of human visual cortex. In recent years, the development of the fMRI technique in non-human primates has enabled neuroscientists to directly compare the topographic organization and functional properties of visual cortical areas across species. These comparative studies have shown striking similarities (‘homologies’) between human and monkey visual cortex. Many visual cortical areas in human can be corresponded to homologous areas in monkey – though detailed cross-species comparisons have also shown specific variations in visual feature selectivity of cortical areas and spatial arrangement of visual areas on the cortical sheet. Comparing cortical structures in human versus monkey provides a framework for generalizing results from invasive neurobiological studies in monkeys to humans. It also provides important clues for understanding the evolution of cerebral cortex in primates. In this symposium, we would like to highlight recent fMRI studies on the organization of visual cortex in human versus monkey. We will have 5 speakers. Each speaker will give a 25-minute talk (including 5 minutes of discussion time). Martin Sereno will introduce the concept of brain homology, elaborate on its importance, and evaluate technical limitations in addressing the homology questions. He will then continue with some examples of cross-species comparison for retinotopic cortical areas. David Van Essen will describe recent progress in applying surface-based analysis and visualization methods that provide a powerful approach for comparisons among primate species, including macaque, chimpanzee, and human. Hauke Kolster will test the homology between visual areas in occipital cortex of human and macaque in terms of topological organization, functional characteristics, and population receptive field sizes. Jonathan Winawer will review different organizational schemes for visual area V4 in human, relative to those in macaque. Reza Rajimehr will compare object-selective cortex (including face and scene areas) in human versus macaque. The symposium will be of interest to visual neuroscientists (faculty and students) and a general audience who will benefit from a series of integrated talks on fundamental yet relatively ignored topic of brain homology.
Evolution, taxonomy, homology, and primate visual areas
Martin Sereno, Department of Cognitive Science, UC San Diego
Evolution involves the repeated branching of lineages, some of which become extinct. The problem of determining the relationship between cortical areas within the brains of surviving branches (e.g., humans, macaques, owl monkeys) is difficult because of: (1) missing evolutionary intermediates, (2) different measurement techniques, (3) body size differences, and (4) duplication, fusion, and reorganization of brain areas. Routine invasive experiments are carried out in very few species (one loris, several New and Old World monkeys). The closest to humans are macaque monkeys. However, the last common ancestor of humans and macaques dates to more than 30 million years ago. Since then, New and Old World monkey brains have evolved independently from ape and human brains, resulting in complex mixes of shared and unique features. Evolutionary biologists are often interested in �shared derived� characters — specializations from a basal condition that are peculiar to a species or grouping of species. These are important for classification (e.g., a brain feature unique to macaque-like monkeys). Evolutionary biologists also distinguish similarities due to inheritance (homology — e.g., MT), from similarities due to parallel or convergent evolution (homoplasy — e.g., layer 4A staining in humans and owl monkey. By contrast with taxonomists, neuroscientists are usually interested in trying to determine which features are conserved across species (whether by inheritance or parallel evolution), indicating that those features may have a basic functional and/or developmental role. The only way to obtain either of these kinds of information is to examine data from multiple species.
Surface-based analyses of human, macaque, and chimpanzee cortical organization
David Van Essen, Department of Anatomy and Neurobiology, Washington University School of Medicine
Human and macaque cortex differ markedly in surface area (nine-fold), in their pattern of convolutions, and in the relationship of cortical areas to these convolutions. Nonetheless, there are numerous similarities and putative homologies in cortical organization revealed by architectonic and other anatomical methods and more recently by noninvasive functional imaging methods. There are also differences in functional organization, particularly in regions of rapid evolutionary expansion in the human lineage. This presentation will highlight recent progress in applying surface-based analysis and visualization methods that provide a powerful general approach for comparisons among primate species, including the macaque, chimpanzee, and human. One major facet involves surface-based atlases that are substrates for increasingly accurate cortical parcellations in each species as well as maps of functional organization revealed using resting-state and task-evoked fMRI. Additional insights into cortical parcellations as well as evolutionary relationships are provided by myelin maps that have been obtained noninvasively in each species. Together, these multiple modalities provide new insights regarding visual cortical organization in each species. Surface-based registration provides a key method for making objective interspecies comparisons, using explicit landmarks that represent known or candidate homologies between areas. Recent algorithmic improvements in landmark-based registration, coupled with refinements in the available set of candidate homologies, provide a fresh perspective on primate cortical evolution and species differences in the pattern of evolutionary expansion.
Comparative mapping of visual areas in the human and macaque occipital cortex
Hauke Kolster, Laboratorium voor Neurofysiologie en Psychofysiologie, Katholieke Universiteit Leuven Medical School
The introduction of functional magnetic resonance imaging (fMRI) as a non-invasive imaging modality has enabled the study of human cortical processes with high spatial specificity and allowed for a direct comparison of the human and the macaque within the same modality. This presentation will focus on the phase-encoded retinotopic mapping technique, which is used to establish parcellations of cortex consisting of distinct visual areas. These parcellations may then be used to test for similarities between the cortical organizations of the two species. Results from ongoing work will be presented with regard to retinotopic organization of the areas as well as their characterizations by functional localizers and population receptive field (pRF) sizes. Recent developments in fMRI methodology, such as improved resolution and stimulus design as well as analytical pRF methods have resulted in higher quality of the retinotopic field maps and revealed visual field-map clusters as new organizational principles in the human and macaque occipital cortex. In addition, measurements of population-average neuronal properties have the potential to establish a direct link between fMRI studies in the human and single cell studies in the monkey. An inter-subject registration algorithm will be presented, which uses a spatial correlation of the retinotopic and the functional test data to directly compare the functional characteristics of a set of putative homologue areas across subjects and species. The results indicate strong similarities between twelve visual areas in occipital cortex of human and macaque in terms of topological organization, functional characteristics and pRF sizes.
The fourth visual area: A question of human and macaque homology
Jonathan Winawer, Psychology Department, Stanford University
The fourth visual area, V4, was identified in rhesus macaque and described in a series of anatomical and functional studies (Zeki 1971, 1978). Because of its critical role in seeing color and form, V4 has remained an area of intense study. The identification of a color-sensitive region on the ventral surface of human visual cortex, anterior to V3, suggested the possible homology between this area, labeled ‘Human V4’ or ‘hV4’ (McKeefry, 1997; Wade, 2002) and macaque V4 (mV4). Both areas are retinotopically organized. Homology is not uniformly accepted because of substantial differences in spatial organization, though these differences have been questioned (Hansen, 2007). MV4 is a split hemifield map, with parts adjacent to the ventral and dorsal portions of the V3 map. In contrast, some groups have reported that hV4 falls wholly on ventral occipital cortex. Over the last 20 years, several organizational schemes have been proposed for hV4 and surrounding maps. In this presentation I review evidence for the different schemes, with emphasis on recent findings showing that an artifact of functional MRI caused by the transverse sinus afflicts measurements of the hV4 map in many (but not all) hemispheres. By focusing on subjects where the hV4 map is relatively remote from the sinus artifact, we show that hV4 can be best described as a single, unbroken map on the ventral surface representing the full contralateral visual hemifield. These results support claims of substantial deviations from homology between human and macaque in the organization of the 4th visual map.
Spatial organization of face and scene areas in human and macaque visual cortex
Reza Rajimehr, McGovern Institute for Brain Research, Massachusetts Institute of Technology
The primate visual cortex has a specialized architecture for processing specific object categories such as faces and scenes. For instance, inferior temporal cortex in macaque contains a network of discrete patches for processing face images. Direct comparison between human and macaque category-selective areas shows that some areas in one species have missing homologues in the other species. Using fMRI, we identified a face-selective region in anterior temporal cortex in human and a scene-selective region in posterior temporal cortex in macaque, which correspond to homologous areas in the other species. A surface-based analysis of cortical maps showed a high degree of similarity in the spatial arrangement of face and scene areas between human and macaque. This suggests that neighborhood relations between functionally-defined cortical areas are evolutionarily conserved – though the topographic relation between the areas and their underlying anatomy (gyral/sulcal pattern) may vary from one species to another.