Hallucinating brains pose for first scans.

A 23-year-old man suffering from schizophrenia sits in a laboratory and hallucinates. He sees disembodied human heads rolling across a vivid backdrop, uttering curses and insults at him.

As this bizarre foray into psychosis begins, the man presses a button that activates brain-imaging equipment. A halo of sensors measures the blood surging and ebbing throughout his brain, enabling scientists to enter the anatomical realm of hallucinations.

Brain scans of this man and five people who frequently hear voices as a result of their schizophrenia have yielded the first direct look at the biological underpinnings of hallucinations, according to a report in the Nov. 9 Nature.

"We now have a map of brain areas involved in the production of hallucinations," contends David A. Silbersweig, a psychiatrist at New York Hospital-Cornell University Medical Center in New York City. "The network of structures that we found is larger than was previously suspected."

Silbersweig and New York Hospital radiologist Emily Stern collaborated with researchers at Hammersmith Hospital in London. Silbersweig and Stern developed several technical innovations that allowed volunteers to activate PET scanners during hallucinations by pressing a button.

A total of 22 to 25 blood flow scans were obtained for each participant, about half during hallucinations and the rest when no voices or visions intruded on their consciousness. The researchers were then able to tag brain areas uniquely linked to hallucinations.

In the five people who heard imaginary voices, blood flow rose markedly in several areas-the hippocampus, parahippocampal gyrus, cingulate gyrus, thalamus, and striatum. These regions, which lie below the brain's outer cortical layer, help to integrate knowledge about one's personal history and emotions with current and past sensations, the scientists assert.

Tissue at the brain's surface involved in hearing also showed blood flow gains during auditory hallucinations.

The man who saw acid-tongued rolling heads displayed comparable, but more widespread, blood flow jumps, notably in cortical areas that integrate sights with sounds, the investigators hold.

Hallucinators also experienced blood flow drops in regions at the front of the brain and in an area that may determine whether voices originate from the self or others. But a structure considered crucial for talking silently to oneself was unaffected by the presence of imaginary voices, Silbersweig notes.

These findings expand on PET data obtained from hallucinators at times when they did not hear voices, he adds (SN: 9/9/95, p.166).

Overactive inner-brain areas that help regulate the chemical messenger dopamine, combined with sluggish frontal-brain structures, may set the stage for hallucinations, Silbersweig theorizes. Inappropriate neural activity in acoustic or visual regions near the back of the brain may then get misinterpreted as a genuine experience, he suggests.