Ever since the Biological Intelligence Research Initiative was formed as an original Beckman Institute research theme, its guiding mission has been to understand human brain function and its role in behavior. These efforts have paid off as BI researchers expand our knowledge of cognition and brain physiology through a unique blend of imaging techniques, theory, and scientific method.
Biological Intelligence researchers have used the facilities at the Biomedical Imaging Center (BIC) and their own innovative imaging systems to understand the workings of the brain. These imaging techniques, used in a variety of research projects, from studying cognitive aging to brain development, are one reason why BI researchers occupy a distinctive place in the field of biological intelligence research. They have formulated groundbreaking theories in the areas of genetics, linguistics, aging, and brain function and physiology.
The Cognitive Neuroimaging Laboratory (CNL) is one setting where this combination of exceptional imaging techniques and leading research topics is yielding remarkable results. CNL co-directors Monica Fabiani and Gabriele Gratton of the Cognitive Neuroscience (CN) group have developed an imaging technology called EROS (Event-Related Optical Signal) that uses a helmet with light fiber sources and optic fiber-bundle detectors. Their prototype was expanded this past year to cover the entire head of subjects, thereby providing researchers data on the full brain – while in action.
“This is really very exciting because it’s like a movie of the brain,” Fabiani said.
Gabriele Gratton has worked on a unique imaging system to study the brain while it is in action.
EROS is a non-invasive neuronal system using the helmet and two to three computers. The system uses diffusion from the light fiber sources on the helmet that are picked up by the detectors; the diffusion differences between active and inactive tissue are observable, and readings are taken. Using the EROS system, Gratton and Fabiani are studying how the brain prepares for tasks, and are able to see what regions of the brain “light up” in preparation for those tasks.
“When a person is given a cue we see them prepare for the upcoming task,” Fabiani said. “When a person needs to switch between one task and another, there are particular areas of the brain that are activated in preparation for the new task.” She said EROS shows that both visual and auditory areas are prepared when needed for the upcoming task, but the area that is not going to be used is actually shut down. “Optically we can look at this action, all of these things as they change over time, unlike fMRI where you have to image everything at the same time.”
Gratton and Fabiani use functional magnetic resonance imaging (fMRI) in their work, but the EROS technique has the advantage of speed, and offers researchers rare insight into how different brain areas interact to solve problems and make decisions.
“Psychologists have been long interested in what mental preparation means. This study is related to finding what brain activity is related to preparation,” Gratton said. “It is part of a whole project whose purpose is to augment cognition, to use physiological measures of brain activity to help people process information.”
Future applications could include physiological sensors being worn by equipment operators or truck drivers to detect whether they are ready for important upcoming tasks or warn them against falling asleep.
CNL researchers have also been studying how physical fitness affects older adults, specifically focusing on less active people who may get less blood flow to the brain. They investigate the relationship of neuronal activity and hemodynamic activity (neurovascular coupling) in these subjects using imaging techniques, and have found that fitness is an important component of mental health, even in older adults.
“In very low-fit older adults we find this relationship is somewhat altered, so blood doesn’t flow as much to active areas of the brain as it does in young adults, or in high-fit older adults,” Fabiani said. “So it’s not just a question of age.”
Maintaining mental health over a lifespan is a key theme that cuts across several disciplines at the Beckman Institute. Denise Park (CN) is co-director with Art Kramer of the Center for Healthy Minds, a research center funded by the National Institute on Aging to study mechanisms for improving cognitive function in later adulthood. Park’s research involves studying the benefits the aging mind can derive from social engagement and skill acquisition. Other Center researchers focus on areas such as cognitive health for older pilots, and engaging older adults in problem-solving tasks.
“A primary goal of the Center is to provide resources to UIUC researchers working on interrelated projects on how to maintain a healthy mind,” Park said. “I think that is a major theme that’s emerging from the Beckman Institute with the various new people who have come on board.”
Kramer, co-chair of the Human Computer Intelligent Interaction Research Initiative, started the Car Talk project to study the effects of cell phone usage on drivers. Kramer works with BI researchers Gary Dell and J.Kathryn Bock of the Cognitive Science (CS) group, and Susan Garnsey of Cognitive Neuroscience (CN) to study the effects of task performance, such as using a cell phone, on test subjects’ driving ability. Using the Institute’s driving simulator, researchers were able to investigate in a novel way this much-discussed topic. The Institute’s driving simulator, with its integrated eye-tracking system, was a perfect investigative tool. Test subjects performed typical driving task such as changing lanes, working the foot pedals, etc. while talking and listening.
“In that way we can compare how much the speech they’re producing or listening to interferes with or changes their driving performance,” Bock said. “We’ve been able to isolate in time where the heavy demand is from listening and talking, and put a number on just how much disruption there is to your driving.”
Their data showed that listening on a cell phone could be just as disruptive as talking on it, or any other task that required a certain amount of attention. While interesting from a news perspective, the research also has meaning for psycholinguists like Bock.
“Most people believe that talking is harder than listening,” Bock said. “But this suggests that listening is just as hard. It’s hard at different times than speaking is, but if you’re really doing it right, if you’re really trying to create in your mind the idea that’s in the speaker’s mind, it’s just as disruptive as what you do to talk. Your mind is occupied with what you are planning to say. And when you’re listening it’s the same thing.”
The study of speech production has implications beyond using cell phones in automobiles.
“Language is one of the most underappreciated things we do and it’s because we do it all the time,” Bock said. “The reason that syntax is important, that structure is important, is that without that there really wouldn’t be ideas. Putting words together – connected speech – conveys the precision of human thought.”
Biological Intelligence researchers are just as willing to challenge traditional theories when it comes to the study of neuroscience.
Jonathan Sweedler of the NeuroTech (NT) group has developed revolutionary sampling protocols for both the mass spectrometry technique called matrix-assisted laser ionization (MALDI) and for capillary electrophoresis. These sampling protocols are extremely useful for analyzing large single cells. Sweedler has used them to study such things as the signaling molecules found within a single neuron and how these molecules are distributed throughout a cell. By looking at the cells of mollusks, Sweedler’s group has found high concentrations of D-Glutamate, a discovery that goes against the conventional wisdom that L-Glutamate is the neurotransmitter used by neurons. This finding has implications for the study of the brain, since the known glutamate receptors respond to the amino acid L-Glutamate – the most important excitatory neurotransmitter in the mammalian brain. Questions asked by this research address whether there are unknown receptors that respond to D-Glutamate or whether a category of L-Glutamate receptor might respond to D-Glutamate. The indications are that the findings in mollusks also apply to the human brain, meaning this research could provide a new perspective on brain function.
“They discovered that mollusks are loaded with this compound that was thought not to exist in the brain and it actually is more powerful than the one that was known to exist,” BI Co-chair Bill Greenough said. “This could seriously rewrite our understanding of the nervous system.”
Bruce Wheeler’s research into the neurophysiology of the brain has also challenged previous thinking on the processes that take place within our heads. Greenough and others went against traditional thinking a decade ago by publishing that protein in brain cells was made from RNA in axons. Wheeler, of the Neurotech group, has taken this now-accepted theory to a new level by using channels that allow only axons to pass through, demonstrating that protein synthesis does in fact take place in axons. This means that neurons are not only transmitters but also are metabolically active parts of brain cells.
Bill Greenough’s research is adding to our knowledge of Fragile X Mental Retardation Syndrome.
Greenough (NT) has focused much of his research on the protein FMRP and the disease Fragile X Mental Retardation Syndrome, which occurs when FMRP is absent. By studying cellular mechanisms the nervous system uses to store information such as memory and unconscious knowledge, Greenough is finding out important facts about Fragile X Mental Retardation Syndrome, the most common genetically inherited form of mental retardation.
“We’ve shown very convincingly that what is principally wrong in the development of the Fragile X MR nervous system is not the initial outgrowth and making of synapses but rather the selective elimination of synapses that gives rise to a patterned adult nervous system normally,” Greenough said.
Greenough Lab member Ivan Jeanne Weiler wrote about these findings in a recent paper published in the Proceedings of the National Academy of Sciences. Weiler said that one of the things deficient at a molecular level in a Fragile X mouse is the ability to make protein in response to stimulation at synapses.
“The problem with Fragile X syndrome is that although the gene was isolated and defined about 10 years ago, it has not been possible to prove exactly what it is that protein does that is so important for the individual (with Fragile X mental retardation),” Weiler said.
Their research demonstrated that a deficiency in FMRP leads to an inability to rapidly create other proteins needed for stimuli response and make the new neuron connections required for normal synapse development.
“What we (reported) in this paper was the Fragile X mice do not have this rapid protein synthesis response,” Weiler said. “The concept now that our lab has developed is that all of our brains have an overproduction of synapses. The connections that are not useful are retracted, the ones that are useful are retained and strengthened.”
For Fragile X patients, both wanted and unwanted synapses are not retracted, leading to a ‘noisy environment’ in their brains. This is an important finding for studies of brain development and mental retardation.
The Greenough lab has quantified this effect and paved the way for future research opportunities and, possibly, treatment. Their research has implications for developing future drug therapies for treatment of mental retardation, in addition to demonstrating some startling results about brain development.
Much of the research taking place within Biological Intelligence will have long-term applications benefiting people in numerous ways. Many of the potential applications are several years in the future, but some BI research projects are already at the engineering phase.
Thomas Anastasio (NT) has been working on a self-aiming security camera system for five years and has now completed a prototype with unique capabilities.
Anastasio’s pioneering research into the superior coliculus region of the brain was the theoretical basis of his system, called the Intelligent Multisensor Fusion camera. This system integrates a video camera, infrared camera and motion detector to follow targets based on probabilities programmed into a computer. The cameras rotate on two axis positioners, and follow a target based on sensory input from detectors pre-programmed to search for things such as heat, motion, or physical features like a face. Just like our brains, the system uses sensors to look for certain data and then responds to that input.
“We think of the brain as the most sophisticated in information processor that’s known,” Anastasio said. “Part of the power of the brain as an information processor is its ability to combine information from multiple sources. Our camera is the only one that combines input from multiple sensors and literally makes an intelligent decision about where it should point.”
An important product of BI research, the Intelligent Hearing Aid, is even further along. Albert Feng’s (NT) previous work on the Intelligent Hearing Aid Project has gone from theory and testing to the licensing stage, and is now close to production.
Speech recognition and language acquisition are focal points of several research projects in the Cognitive Science group. Aphasia is the loss of language ability due to brain damage, and researchers in CS are among the leaders in developing ways to understand this disorder. Dell has set up a Web-based computational model that goes beyond other methods in this field by allowing information from tests of aphasia patients to be entered in as data, and then have the computational model applied to that data. Dell said this permits a way of “diagnosing” the patient.
“Specifically, the model then assigns a value to the patients' semantic abilities, and another number to their phonological abilities,” Dell said. “It also describes how consistent the patient is or is not with the model's assumptions.”
Dell said researchers working any place in the world who have tested an aphasia patient could apply the model to their results. Dell has applied the model to data from more than 100 patients from a Philadelphia study.
This research could lead to therapies for patients with aphasia that will improve their speech abilities. It also helps to understand the structure of language, Dell said. “By studying the loss of language we develop theories or characterizations of what language is like when it’s broken down.”
Dell has also applied a computational model to a long-running collaboration with Bock and Neal Cohen (CN) that is concluding with some interesting results. Their research showed that even people who have suffered brain damage due to strokes and suffered severe memory loss were able to learn and remember language constructs.
“Even in older adults, even in amnesiacs, even in people who have lost the ability to remember what happened to them in the course of their lives, they continue to learn these things about language,” Bock said. “And it’s something that lasts for a fairly long time.”
Dell, Jennifer Cole (CS), and Cynthia Fisher (CS) are investigating how children and adults learn phonological patterns in language. Understanding how this phonological processing goes on in the brain can shed light on how the human brain acquires new information about language structure. Their research is unique in terms of pairing adults and children, the computational models resulting from the work, and the teamwork between linguists and psychologists.
Jennifer Cole focuses on how spoken language is processed in the brain.
Cole said they are interested in learning how the pronunciation and sound of a word is encoded in the brain, as well as how an individual learns about the sound patterns of their language, and what kind of information in the linguistic environment is guiding the individual’s knowledge.
“We’re trying to come up with some pretty concrete models about how the phonological information about words is encoded in the brain and how we build up those encodings, that phonological knowledge, through our experience of speaking and listening to our native language,” Cole said.
Some of the work comparing adults and children has produced some interesting results.
“One of the things we’re finding is that adults also continue to learn and that our knowledge of language isn’t stagnant,” Cole said. “Your knowledge of your native language continues to unfold over your lifetime.”
Cole and Mark Hasegawa-Johnson of HCII are in the third year of a collaboration that started as a University of Illinois Critical Research Initiative on new approaches to speech recognition. They are focusing on integrating prosody (the rhythm and intonation of spoken language) into speech recognizers (algorithmic models run in software) and, someday, speech recognition devices. This past year they have focused on disfluency, a common element of speech that has proven difficult for machines to comprehend. Cole said they are trying to develop tools that will automatically detect regions in speech where speakers are being disfluent with ‘ums’ and ‘ahs’ or repetitions or false starts.
“Linguists have known about these patterns speakers produce,” Cole said. “The trick has been to develop computational models, signal processing tools, and machine learning tools to successfully recognize these features in speech.”
Cole said they have had some initial success and are now trying to discover the information in the acoustic signal that serves as a landmark for those disfluencies.
The possibilities in terms of technologies such as improved speech recognition devices that come from Biological Intelligence research seem unlimited. Just as important, BI researchers are finding that the brain’s capabilities are greater than what was previously thought.
Park is involved in a new project that early on is demonstrating some interesting findings about culture and the plasticity of the brain. Park and Singapore neurologist Michael Chee are leading a global study on how cultural influences may shape the brain and as a result, cognition, in ways distinctive to that culture. Park said the early phase of the study, which focuses on neural correlates of cultural experiences, demonstrates differences in neural pathways used by older Asians compared to Americans when it comes to recognizing complex scenes.
The study uses magnetic resonance imaging, and compares how East Asians and Westerners process scene information from a picture. They are finding differences in the way objects and backgrounds in pictures are processed with respect to age and culture.
“We are finding that older East Asians activate neural pathways associated with background scenes when they look at pictures, and Americans activate pathways associated with individual objects.” Park said. “The differences become very pronounced in older adults.”
The study is not only innovative, but also is unique.
“We hope to ultimately spend a number of years testing the ways that cognitive function leads to differences in neural function and also to do the first cross-cultural study of the aging brain,” Park said. “Nobody has done that before.”
So far, the early results of the study are showing the same findings about the plasticity of the brain that have come out of other BI research.
“Both the Center research and culture study are based on the theory that the brain changes as a result of experience,” Park said. “Cultural experiences change your brain, engaging experiences change your brain. I actually think this work fits thematically within a broader picture of interdisciplinary research that is occurring at the Beckman.”
And that picture is the growing awareness of the plasticity of the brain. Research from a wide variety of projects in Biological Intelligence is showing how this malleability of the brain and its capacity to acquire new information, even later in life, will change how we view everything from mental health in older adulthood to developing drug therapies for treating disease.