Spreading consciousness: awareness goes global in the brain.

Brain-scan images are iridescent icons of today's science of the mind. Where questionnaires, interviews, and observations of behavior once reigned supreme, fancy machines hooked up to fancier computers now create portraits of brains at work. As volunteers complete memory tests or some other mental task, molecular changes inside their skulls get transformed into images sporting multicolored splotches signifying pockets of heightened brain activity. Scientists usually regard these gaudy neural patches as products of specialized brain structures that coordinate the mental process under study.

From a statistical perspective, though, there's more than one way to skin a brain scan. In fact, a fresh treatment of imaging data indicates that at least one crucial aspect of thought recruits the entire brain, not just a few of its parts, says philosopher Dan Lloyd of Trinity College in Hartford, Conn. His unique statistical strategy has drawn kudos for ingenuity from many neuroscientists, even though most still favor standard procedures for turning raw data into brain images.

Philosophical ideas about the nature of consciousness inspired Lloyd's analytical innovation. Conscious thought is tough to define, much less to study. Still, a burgeoning number of imaging studies link bursts of activity in several far-flung brain areas to a person's awareness of specific objects or features.

That, at least, is the implication of standard analyses that lump together brain data collected from groups of people as they perform visual tasks. These methods then prune away neural responses presumed to be extraneous to the tasks and finally highlight discrete brain regions that exhibited dramatic surges of activity.

By its nature, however, consciousness eludes efforts of brain imagers to corral it into neural pens, Lloyd argues. There's no seat of consciousness in the brain, in his view. The whole brain mediates awareness as a person, for example, gazes at an object or a scene.

In this scenario, consciousness requires a brain-based sense of time, melding an awareness of what one has just perceived, what one currently perceives, and what one expects to perceive in the coming moments. An ongoing mingling of past, present, and future experiences makes our perceptions seem to flow past us.

A statistical reanalysis of four sets of functional magnetic resonance imaging (fMRI) data supports this theory, Lloyd reports in the Aug. 15 Journal of Cognitive Neuroscience.

"There's a global brain state that's constantly in flux and that creates conscious experiences," Lloyd says. "That's why you can't step in the same stream of consciousness twice."

BLIND DATA Cognitive neuroscientists don't simply snap pictures of neurons' private doings like paparazzi peering into a celebrity's bedroom with a telephoto lens. Instead, they choose particular statistical methods to transform brain data into images that portray what's going on in small, dispersed patches of brain tissue.

Consider fMRI. This technology applies powerful magnetic signals to the brain and then detects changes in the relative amount of blood flowing in different parts of the brain. In fMRI, it takes several seconds or more to generate a single brain image as a person addresses a mental task.

Researchers typically calculate the average blood-flow rate throughout the brain for a group of volunteers during an experimental procedure, such as reading actual words, and during at least one control procedure, such as reading nonsense words. Brain activity in the control condition is then subtracted from that in the experimental condition, leaving--in theory--responses that are unique to the latter procedure.

Lloyd's alternative approach proceeds from assumptions about consciousness proposed more than 60 years ago by the philosopher Edmund Husserl. According to Husserl, each person's conscious states distinctively refer to objects in the world and fill in actual sensations with extra information, such as seeing an entire piece of white paper when it's partly covered by a person's hand. He also held that conscious perceptions are always in flux bemuse they're a blend of past perceptions, current experiences, and a sense of what's coming up.

To test these possibilities, Lloyd turned to the fMRI Data Center at Dartmouth College in Hanover, N.H. Launched in 2000, the center provides researchers access to data generated in all studies published in the Journal of Cognitive Neuroscience. The center last year sponsored a research award for investigators who had used its stored data in new studies. Lloyd won the $5,000 award over more than 20 other competitors.

Lloyd probed data from four studies published in 2000. In one of them, a team headed by Eliot Hazeltine of the NASA Ames Research Center in Mountain View, Calif., examined areas of neural activity as people responded to a circle's changing color. In a second report, Alumit Ishai of the National Institute of Mental Health in Bethesda, Md., and her coworkers identified distinctive brain regions activated as volunteers looked at pictures of faces, houses, and chairs.

A third project, directed by Andrea Mechelli of University College London, tracked changes in brain activity as people silently read words slowly or rapidly. Finally, Bradley R. Postle of the University of Wisconsin-Madison led a study of neural areas implicated in both visual memory and eye movements.

In all their analyses, the teams combined data from groups of volunteers to construct images of brain activity.

Flouting convention, Lloyd separately examined brain data for each of the 27 individuals in the four experiments. He measured the extent to which blood flow changed in each digitized dot of a participant's brain during test sessions.

Activity throughout each brain changed gradually over time, so that any particular image looked much like the images immediately before and after it but showed increasing disparities with images coming earlier and later in the session. Brain activity didn't suddenly spike as a person completed a particular task or made a visual choice.

This result is consistent with Husserl's contention that individuals formulate unique conscious perceptions of the world that are in constant flux, Lloyd asserts.

Moreover, he finds that simple learning systems within computers--known as neural networks--can be trained to use data from a series of any volunteer's brain images to create a close copy of both the immediately preceding and the immediately following image. According to Lloyd, this finding indicates that current brain states contain information that reflects past and future brain states, in concert with Husserl's emphasis on time as a key component of consciousness.

"We're going to have to shift gears radically to study consciousness effectively," Lloyd says. "Time perception is critical, and it appears to be a global property of each individual's brain."

GLOBAL PROMISE Those who take radical liberties with other scientists' data usually get about as much respect for their efforts as does a deposed Enron executive giving an ethics lecture. However, Lloyd's study gets surprisingly positive reviews from the researchers whose data he borrowed.

"The findings potentially offer a novel and groundbreaking picture of the workings of the human brain," comments NASA's Hazeltine. Further work needs to explore whether other theories can account for global patterns of brain activity better than Lloyd's does, he adds.

For instance, Lloyd's results may not reflect consciousness-related properties of the brain, remarks Wisconsin's Postle. Instead, gradual changes in brain activity may simply reflect a person's getting used to sitting in a scanning device and performing a task. Extraneous data generated by fMRI scanners and not statistically controlled for by Lloyd could also have contributed to his findings, according to Postle.

Still, he says, "Lloyd's study suggests an approach to consciousness that marks a promising departure from the status quo."

Ishai and Mechelli also refrain from making any firm conclusions based on Lloyd's report. Yet both see it as a demonstration that data sharing can stimulate creative new approaches to brain-imaging research.

Lloyd shows that philosophers potentially have much to teach cognitive neuroscientists about how to study the mind, comments John D. Van Horn, who organized the Dartmouth Brain Imaging Center's 2001 research award.

A new generation of philosophers of the mind stands ready to dive into the neuroseience data pool, says philosopher Daniel C. Dennett of Tufts University in Medford, Mass. Lloyd's "bold" analysis may be the first of many, in his view.

"We're drowning in data from standard fMRI studies," Dennett adds. "New theoretical perspectives about the brain are needed."

BLURRY IMAGES Concern has grown in the past few years that brain-imaging devices as currently deployed can't penetrate the mysteries of conscious thought. In a paper on this topic--titled "Can functional brain imaging discover consciousness?"--philosopher Antti Revonsuo of the University of Turku in Finland concluded that it's possible, but not with current methods. The work was published in the March 2001 Journal of Consciousness Studies.

Several problems loom large, in Revonsuo's opinion. First, fMRI generates images too slowly to capture the breakneck brain changes associated with consciousness. Second, fMRI and positron emission tomography, or PET, tap into changes in blood flow rather than neural activity directly involved in conscious thought. Third, as stressed by Lloyd, brain images reveal only that, on average, certain areas become more active during experimental conditions. This offers no insight into the function of those areas.

Broadly distributed, lightning-quick changes in the electrical properties of thousands or even millions of neurons probably foster consciousness, Revonsuo theorizes.

A few researchers have pursued this possibility. Measurements of electrical patterns in the brain reveal widespread synchronized activity during conscious perceptions, but not when people perceive information without being aware of it (SN: 2/20/99, p. 122). Such findings indicate that conscious experience arises thanks to dusters of neurons throughout the brain that cooperate to extract meaning from sensations in fractions of a second, according to a theory proposed by Giulio Tononi of the University of Wisconsin-Madison and Gerald M. Edelman of the Neurosciences Institute in San Diego.

In their view, for example, an individual's conscious feeling of, say, redness reflects a unified neural process that incorporates increased activity in red-selective neurons with readings from many other neuronal groups. These include those selective for green and blue, for visual motion and shape, and for emotions.

Walter J. Freeman of the University of California, Berkeley takes a similar approach (SN: 11/2/96, p. 280). Using electrodes surgically placed on the brain, Freeman and his coworkers find that cats, gerbils, rabbits, and rats display fleeting patterns of activity across large swaths of tissue as they learn to categorize sounds or smells.

In a 2001 study, a group of Mongolian gerbils learned to discriminate between rising and falling tones. As gerbils then heard new examples of rising and falling tones and reacted appropriately, they exhibited split-second changes in patterns of electrical activity across the entire section of their brains devoted to sound perception. Such patterns, unique to the individual animals, reflected an ongoing brain process of generating and testing hypotheses about what's going to happen based on past experiences, Freeman proposes.

Dynamic activity patterns in the brain that are associated with conscious recognition of learned categories transcend the small, colored patches in fMRI images, he argues. By averaging data from many brains and then subtracting responses to control conditions from those to experimental conditions, fMRI researchers artificially create islands of brain activity for whatever they happen to be studying, in Freeman's view.

Most cognitive neuroscientists disagree. Lloyd himself sees value in exploring both standard and alternative fMRI analyses of memory and other mental functions. He suspects, though, that ultimately consciousness will yield only to global studies of individual brains.

As Husserl would undoubtedly add, only time will tell.