2018: FREiWALD


Pioneering work uncovers how the brain processes faces

Faces are our primary source for recognizing people and reading their emotional and mental states. The brain’s visual system extracts social meaning from faces and activates other brain circuits to generate emotional reactions, trigger memories, direct attention, and guide actions toward others. For the past 15 years, Winrich Freiwald, a professor at The Rockefeller University, has been on a mission to understand the neural mechanisms and computations of the face-processing network, thereby uncovering fundamental principles of brain organization.

Using functional magnetic resonance imaging (fMRI), Freiwald co-discovered a specialized neural machinery for face processing located in the temporal and frontal lobes of the brain. By combining fMRI with electrophysiological techniques, he and his colleagues further showed that this machinery is composed of a small network of a fixed number of face-selective regions, called face patches, each dedicated to a different aspect of face processing and all closely connected with each other.

“We discovered that the brain processes faces through a dedicated machine,” Freiwald said. “This machine is a small network of brain areas, each devoted to the analysis of faces, and each generating a unique representation of faces from those the other areas provide. It is a beautiful system, and it helps us mechanistically understand a problem that long seemed too difficult for neuroscience to crack.”

Because a dedicated circuit exists for processing them, faces offer a unique opportunity to study object recognition. Likewise, as potent stimuli for attention, emotions, memories, and thoughts, faces provide a powerful means to study social cognition. Shedding light on the social brain circuits that implement these complex functions may aid in understanding disturbances of social behavior in diseases such as autism.

For his seminal contributions to the field of visual neuroscience, Freiwald has been named the recipient of the 2018 Golden Brain Award from the Berkeley, California-based Minerva Foundation. The award, now in its 34th year, recognizes outstanding contributions in vision and brain research. Freiwald was honored in a private ceremony held on December 14th in New York City.

“The Golden Brain Award has recognized some of the most outstanding researchers in systems and cognitive neurosciences,” Freiwald said. “Many of the honorees are scientific heroes of mine who I will always admire. Being honored among them is most humbling to me.”

Entering a burgeoning new field

A native of Oldenburg, Germany, Freiwald studied biology and mathematics as an undergraduate student at the Georg-August University of Göttingen and the Eberhard Karls University of Tübingen. He then performed his graduate work at the Max Planck Institute for Brain Research in Frankfurt, where he started working on visual object recognition with Max Planck Director Wolf Singer.

After receiving his Ph.D. from the University of Tübingen in 1998, Freiwald joined the Institute for Brain Research at the University of Bremen as a lecturer. Starting in 2001, he worked as a postdoctoral fellow at the Massachusetts Institute of Technology (MIT), Massachusetts General Hospital, Harvard Medical School and the Hanse Institute for Advanced Study in Delmenhorst, Germany. 

Freiwald was introduced to the field of cognitive neuroscience as a postdoctoral fellow in the lab of former Golden Brain Award recipient Nancy Kanwisher, the Walter A. Rosenblith Professor of Cognitive Neuroscience in the Department of Brain and Cognitive Sciences at MIT. “It was an exciting and incredibly stimulating time, a time of many new discoveries in a burgeoning new field,” Freiwald said.

Later, as a postdoctoral fellow in the lab of Margaret Livingstone, the Takeda Professor of Neurobiology at Harvard Medical School, Freiwald learned just how much he loves doing experiments. He began to use fMRI to discover face areas in the brain when he joined forces with former Golden Brain Award recipient Doris Tsao, then a graduate student in the Livingstone lab and now a professor of Biology at the California Institute of Technology.

Freiwald was director of a brain-imaging group at the Centers for Advanced Imaging and Cognitive Science at the University of Bremen in Germany, from 2004 to 2008, and then a visiting associate at the California Institute of Technology in 2009. During that time, he continued to investigate the organization and main properties of the face system. Since joining the faculty at The Rockefeller University in 2009, Freiwald has focused on elucidating the inner workings of this system, from the level of individual cells to interactions among brain areas.

“He flies high over the brain to overview it as a whole, yet at the same time sees the brain in great detail with extremely high resolution,” said Atsushi Iriki, a former Golden Brain Award recipient and head of the Laboratory for Symbolic Cognitive Development at the RIKEN Center for Biosystems Dynamics Research. “His unique ways of designing research by integrating global and elementary viewpoints made this possible. His research will clearly advance the future direction of this research field over coming generations, during which he is expected to be a leader and a real giant.”

Discovering face patches

In a seminal Science study published in 2006, Freiwald, Tsao and their collaborators combined fMRI with single-cell recordings to discover a cortical region consisting almost entirely of face-selective cells, which respond more strongly to faces than to other types of objects. In a study published in Science two years later, they went on to report a unified, interconnected system of cortical areas, called patches, dedicated to processing faces. Electrical stimulation of face patches produced strong activation in other face patches, as revealed by fMRI.

In 2009, the researchers reported in Nature Neuroscience that face-patch neurons detect and differentiate faces using both part-based and holistic strategies. Face cells appear to be feature detectors, working best in the holistic construct of a face. The cells detect distinct constellations of face parts and are tuned to the geometry of facial features. Intriguingly, face cells tend to be tuned to extremes of face parameters, possibly explaining why caricatures can be so effective. Essentially, face cells span a face-feature space, a fundamental property of human face recognition.

As Freiwald reported in another study subsequently published in Science, some of the face-patch neurons respond to specific viewpoints of faces, while others do not. At the top of the face-patch hierarchy, the cells respond selectively to individual faces across transformations in size, position, and head orientation, suggesting that they may play a key role in recognizing faces across different viewing conditions.

“The core face-processing system exhibits three main organizing principles: concentration of cells selective for the same object category into modules, spatial separation of modules with different functional specializations, and tight coupling between face areas,” Freiwald said.

Mapping the social brain and attention networks

By studying how the face-processing system is embedded in the brain, the Freiwald lab is determining the links between face recognition and social cognition, thereby helping to understand disturbances of such behaviors in disease. Because it is so closely devoted to the processing of the diverse social cues the face conveys, the face-processing network is an ideal entry point to investigate the neural mechanisms of social cognition.

“This work will allow us not only to understand how faces are processed, but how neurons interact to generate the key characteristics of social cognition,” Freiwald said. “With this work, we also aim to lay the foundation for future work in autism models to understand the neural mechanisms underlying the alterations of social-information processing in this condition.”

Toward this goal, the Freiwald lab showed how the face-processing network is embedded in the social brain, revealing connections with other brain regions involved in emotion processing, memory, and cognitive functions. In a study published in Science in 2017, the Freiwald lab discovered two face areas selective for familiar faces, a link between face perception and person knowledge. In the same year, in another study published in Science, the Freiwald lab discovered a large brain network dedicated to observing and processing social interactions, a potential foundation for high-level social cognition. Furthermore, the lab discovered the functional brain circuit controlling facial movement, and its link to face perception and social cognition circuits.

“As face-to-face social conversations have made humans an especially successful species among other animals, understanding its fundamental brain mechanism is key to understanding the nature of human civilization,” Iriki said.

The Freiwald lab also uses fMRI to characterize the entire network of areas involved in attention and to identify its connections and functional properties. The group discovered one network that includes a new cortical area not previously suspected to be involved in attention, located in the temporal lobe right in between two face areas. It is connected to other areas involved in attention, but may play a different role in focusing attention. “This discovery was entirely unexpected and is still so fascinating to me. I think it is a breakthrough finding in the domain of attention,” Freiwald said.

Current research in the Freiwald lab is aimed at uncovering what makes us intelligent. “The approach we have been using, the integration of fMRI with targeted investigations of functions within fMRI-identified regions and computational modeling, has been critical to uncover this deepest mystery of our brains. As our discovery of new areas for social cognition shows, it is possible for us now to solve the problem. Our approach is impactful for cognitive and systems neurosciences at large,” Freiwald said. “In future studies, I will continue to show the power of this approach in the domains of face recognition, attentional control, and social cognition.”