Four Eagles games, four concussions - the latest two suffered by wide receiver Riley Cooper and cornerback Asante Samuel on Sunday. Coach Andy Reid said Monday that their injuries "seem to be mild," but that word seem is telling.

From a scientific perspective, no one really knows. Physicians can see when someone's outward symptoms have returned to normal, as most do within a few weeks. But there is no lab test to measure internal damage from a concussion and no medicine to treat it.

"Absolutely nobody knows when it's safe to go back in," said Douglas H. Smith, director of the University of Pennsylvania's Center for Brain Injury and Repair. "If anyone says one week is good, three weeks is good, they're not basing it on facts."

The scientific study of concussions, sometimes called mild traumatic brain injuries (TBI), lags behind research on many other ailments - a failing attributed to a lack of funding and public pressure, among other reasons. Smith and a growing number of others are striving to change that, investigating techniques such as advanced brain scans, genetic screens, and the analysis of biomarkers - small amounts of certain proteins that may leak into the blood after a brain injury.

They want to know if there has been subtle damage inside brain cells - the sort of thing that might lead to chronic problems years later.

The quest has taken on new urgency with the prevalence of soldiers who suffer head injuries from the jolt of an explosion, experiencing cognitive deficits despite the lack of external wounds. Concern over the issue has filtered through to civilians as well, from professional athletes down to neighborhood leagues.

The sports world has paid particular attention to an ongoing study at Boston University, where researchers have analyzed the brains of deceased football players, identifying numerous cases of chronic traumatic encephalopathy - a degenerative disease marked by "tangles" in the brain.

But the goal is to peer inside the body while the patient is still alive. An ideal solution: a simple blood test that would allow the diagnosis of concussions within minutes, on the playing field or even in a war zone.

One biotech firm working to develop such a test is called Banyan Biomarkers, founded by scientists from the University of Florida. The Department of Defense plans to collaborate with the firm on a phase-three trial of its biomarkers, said Col. Dallas C. Hack, who is based at Fort Detrick, Md., and directs the Combat Casualty Care Research Program.

The key is to make sure that such proteins are specific to brain injury, and not something that turns up in the blood after injuries elsewhere in the body. Kevin Wang, a Banyan cofounder, said he and his colleagues had identified several good candidates, including one protein normally found in the cytoplasm of brain cells.

"We're just taking what Mother Nature gives us," Wang said.

Biomarkers are under scrutiny at Penn, too, as part of a joint study with the Baylor College of Medicine. That study also includes research on a kind of advanced MRI called diffusion tensor imaging (DTI), both in humans and pigs.

A traditional MRI does an excellent job of distinguishing among different kinds of tissue, in part by measuring the density of protons in bodily fluid. But it's not useful for looking at anomalies within a given type of tissue of uniform density, such as white matter - the fatty, insulating tissue that covers axons in brain cells.

The type of cellular injury that is thought to be associated with concussions, in other words, can't be seen in a regular MRI.

That's where DTI may be useful, allowing the measurement of protons contained in water molecules. Scientists can detect how these protons are moving inside an axon, the long, skinny portion of a neuron. In a healthy axon, the water protons are more free to travel along its length, said Elias Melhem, a neuroradiologist at Penn. In an axon that has been stretched, suffering internal damage and even swelling, the paths of the water protons are more random, he said - a quality that can be quantified with sophisticated mathematics.

Lots of work remains to determine how the results should be interpreted, said Alisa Gean, a neuroradiologist at the University of California, San Francisco. In theory, DTI measures the structural integrity of axons, but it's not clear yet how the measurements correspond to actual brain function, she said.

"While it does hold great promise, it's not ready for prime time," she said. Other kinds of advanced imaging are being studied as well, in some cases revealing internal abnormalities even after the patient reports feeling back to normal, Gean said.

To get a better look at individual axons, Smith's lab also puts rat brain cells on a silicon membrane and then stretches them by administering puffs of air. The amount of stretching, and the speed at which it occurs, is akin to what a human brain would suffer in a mild traumatic brain injury, Smith said, though he doesn't like the adjective mild because the consequences can be grave.

The results are not pretty. Looking at the stretched cells under a microscope, the scientists can see breakage in internal structures called microtubules, which Smith likens to train tracks that carry crucial proteins along the length of an axon. This can lead to a pileup of proteins, causing swelling and even disintegration of the cell.

Even if there is no internal breakage, stretching an axon too quickly leads to chemical changes that appear to make the cell more susceptible to damage from a repeat stretch a day later, he and colleagues have reported.

"These types of changes may underlie the suspected 'period of vulnerability' for athletes suffering concussion that is the center of debate about when they should return to the game," Smith said.

Another challenge is studying the sort of axonal stretching thought to occur during the blast wave from an explosion. The amount of stretching is much less than when a football player's head is slammed into the turf, but the rate of stretching is faster, due to the high frequency of the blast wave, Smith said.

Injuries can occur even when the soldier suffers no physical impact or external injury. But how to measure these unseen impacts?

With engineering colleagues Shu Yang and Kacy Cullen, Smith has developed a small, wearable sensor that will change color when subjected to strain at the high frequency of a blast wave. The sensor is made of a crystalline material, which responds by collapsing at a certain frequency - a little bit like an opera singer shattering a crystal glass. The color of the sensor changes because the collapsed structure refracts light differently.

These various lines of research are welcomed by neurologists, said Jeffrey S. Kutcher, chair of the American Academy of Neurology's Sports Neurology Section. Kutcher is involved in re-writing the academy's concussion guidelines to reflect some of the latest studies.

For example, it used to be considered medically acceptable for athletes to return to action if they had mild symptoms that cleared up within 15 minutes of being hit, said Kutcher, who is also director of Michigan NeuroSport, a medical and research team at the University of Michigan. But symptoms can arise or worsen up to several hours later, he said.

Much more science is needed, he and other brain-injury experts said. Obtaining consistent research data is difficult because an identical impact can cause a range of responses in different people, Kutcher said.

"Concussion is not like an ACL injury. Everybody's knee is essentially the same. We know what it does, we know what it should do," allowing for a standard treatment protocol, Kutcher said. "Brains are way too diverse for that."

Then there's the problem that concussions have historically been underreported, because of the old attitude that athletes should shake it off and get back on the field - an approach that neurologists say is dangerous. Estimates of the number of concussions vary widely, but tend to place it well above one million each year.

One of the biggest risk factors for concussion is having already had a concussion. But it is not entirely clear what that means.

The first concussion might have weakened those people and made them more susceptible to repeat injury. Or perhaps they were more prone to having concussions to begin with, either because of the nature of their brains or their occupation. Outward symptoms are meaningful, but Kutcher said indicators from inside the head are desperately needed. Hack, the head of the military research program, agreed.

"This is truly an under-addressed area of medicine, one of the last frontiers of medicine that we have not done a good job on as a medical community," Hack said. "We're going to change medicine with the work we're doing."

Contact staff writer Tom Avril at 215-854-2430 or