A large-scale study published in the New England Journal of Medicine has sparked a flurry of controversy among anesthesiologists. According to the findings, a commonly used device designed to prevent anesthesia awareness–the rare event when a patient is actually conscious during surgery–was largely ineffective.

The findings highlight just how little is known about the neural changes that underlie anesthesia. “The challenge is that we don’t understand the physiology and pharmacology underlying memory blocking by anesthetics,” says Beverly Orser, an anesthesiologist and scientist at the University of Toronto, who wrote an editorial accompanying the piece. “If we understood the circuits and brain regions involved in complex memory formation, we’d be in a better position to develop these monitors.”

Emery Brown, an anesthesiologist and neuroscientist at Massachusetts General Hospital, aims to do just that. Brown and his colleagues are using both brain imaging of human volunteers and, in animals, electrophysiology approaches–which more directly measure brain activity–to gain a deeper understanding of anesthesia. Preliminary research from his lab suggests that measuring activity at the surface of the brain may not be a reliable indicator of what’s going on deeper down, where the memory circuitry may still be functioning–and forming frightening recollections of a particular surgery.

Every year, more than 20 million people in North America undergo general anesthesia–a combination of drugs that sedate patients, paralyze their muscles, and block perception of pain. The cocktail is carefully titered to each individual and each surgery, with the aim of maintaining the patient’s crucial functions, such as heart rate and blood pressure, while keeping her blissfully unaware of the procedure.

A small number of those who get general anesthesia–about 0.1 to 0.2 percent–will experience awareness, which ranges from relatively innocuous incidents, such as later remembering a conversation between surgeons and nurses, to reports of excruciating pain while completely paralyzed. While it’s not exactly clear what triggers anesthesia awareness, an insufficient amount of drugs that quiet brain areas involved in learning and memory is thought to be part of the problem.

As recognition of the problem of anesthesia awareness has grown in recent years, so has the market for devices designed to prevent it. Several types of monitors are now commercially available. They are based on a simple concept: that anesthesia drugs quiet the cortex in a predictable manner that can be measured with electroencephalography (EEG), a technology that measures electrical activity on the surface of the head. The frequency of brain waves spikes briefly as the patient is lulled into unconsciousness, and then it slows. The devices convert EEG patterns into a single number that indicates a patient’s level of awareness, allowing physicians to administer more drugs if needed.

But Brown and others argue that devices like this give only a rudimentary measure of what’s happening in the brain. “If it’s slow, we think it’s okay to operate; if it’s fast, we think they’re waking up,” says Brown. “That’s all we’re doing.”

Brown and his colleagues are using newly developed technology that allows them to study EEG waves while a patient is simultaneously having his brain imaged with functional magnetic brain imaging, an indirect measure of brain activity that is more spatially precise than EEG. Preliminary results show that some brain areas actually become more active during the course of anesthesia. It’s not surprising that a broad-acting drug, which inactivates brain areas that are normally involved in selectively inhibiting brain activity, leads other areas become more active, says Brown. “This is the type of information we really need,” he says.

In corresponding experiments conducted on rodents, scientists used arrays of electrodes to directly measure activity in different parts of the brain. Researchers directed by Matt Wilson, a professor of brain and cognitive sciences MIT who collaborates with Brown, found that rodents that had been given an increasing dose of an anesthetic showed characteristic changes in the rhythm of brain activity in the cortex. But activity in the hippocampus, a brain area crucial in learning and memory, remained unchanged.

“If the signature [measured via EEG] is coming from the cortex, it’s not telling us what the deeper brain structures are doing, such as the arousal system, the brain stem, the amygdala, and the hippocampus,” says Brown. “If EEG cannot tell you about those structures, it’s not telling you about key systems.”