t first it seemed like nothing more than the worst headache she’d ever had.

So Karla Pérez—22 years old, the mother of three-year-old Genesis, and five months pregnant—went into her mother’s room to lie down, hoping it would pass. But the pain got worse, and as she vomited off the side of the bed, she told her younger brother to call 911.

It was not quite midnight on Sunday, February 8, 2015. The ambulance raced Pérez from her home in Waterloo, Nebraska, to Methodist Women’s Hospital in Omaha. She began to lose consciousness in the emergency room, and doctors put a tube down her throat to keep oxygen flowing to her fetus. They ordered a CT scan, and there it was: a massive brain bleed creating severe pressure in her skull.

She had suffered a stroke, but amazingly her fetus was doing fine, the heartbeat strong and steady as if nothing were wrong. Neurologists did another CT scan at about two in the morning, and their worst fears were confirmed: Pérez’s brain had become so swollen that the whole brain stem had pushed out through a small opening at the base of her skull.

“When they saw that,” says Tifany Somer-Shely, the obstetrician who’d cared for Pérez through her pregnancy with Genesis and with this baby too, “they knew for sure that it wasn’t going to end well.”

Pérez had landed at the ragged border between life and death, with a brain that had ceased functioning and would never recover—in other words, it was dead—and a body that could be sustained mechanically, in this case for one reason only: to nurture her 22-week-old fetus until he was big enough to manage on his own. This borderland is becoming increasingly populated, as scientists explore how our existence is not a toggle—“on” for alive, “off” for dead—but a dimmer switch that can move through various shades between white and black. In the gray zone, death isn’t necessarily permanent, life can be hard to define, and some people cross over that great divide and return—sometimes describing in precise detail what they saw on the other side.

Death is “a process, not a moment,” writes critical-care physician Sam Parnia in his book Erasing Death. It’s a whole-body stroke, in which the heart stops beating but the organs don’t die immediately. In fact, he writes, they might hang on intact for quite a while, which means that “for a significant period of time after death, death is in fact fully reversible.”

How can death, the very essence of forever, be reversible? What is the nature of consciousness during that transition through the gray zone? A growing number of scientists are wrestling with such vexing questions.

In Seattle biologist Mark Roth experiments with putting animals into a chemically induced suspended animation, mixing up solutions to lower heartbeat and metabolism to near-hibernation levels. His goal is to make human patients who are having heart attacks “a little bit immortal” until they can get past the medical crisis that brought them to the brink of death.

In Baltimore and Pittsburgh trauma teams led by surgeon Sam Tisherman are conducting clinical trials in which gunshot and stabbing victims have their body temperature lowered in order to slow bleeding long enough for surgeons to close up their wounds. The medical teams are using supercooling to do what Roth wants to do with chemicals—kill their patients, temporarily, in order to save their lives.

In Arizona cryonics experts maintain more than 130 dead clients in a frozen state that’s another kind of limbo. Their hope is that sometime in the distant future, maybe centuries from now, these clients will be thawed and revived, technology having advanced to the point where they can be cured of whatever killed them.

In India neuroscientist Richard Davidson studies Buddhist monks in a state called thukdam, in which biological signs of life have ceased yet the body appears fresh and intact for a week or more. Davidson’s goal is to see if he can detect any brain activity in these monks, hoping to learn what, if anything, happens to the mind after circulation stops.

And in New York, Parnia spreads the gospel of sustained resuscitation. He says CPR works better than people realize and that under proper conditions—when the body temperature is lowered, chest compression is regulated for depth and tempo, and oxygen is reintroduced slowly to avoid injuring tissue—some patients can be brought back from the dead after hours without a heartbeat, often with no long-term consequences. Now he’s investigating one of the most mysterious aspects of crossing over: why so many people in cardiac arrest report out-of-body or near-death experiences, and what those sensations might reveal about the nature of this limbo zone and about death itself.

Oxygen plays a paradoxical role along the life-death border, according to Roth, of Seattle’s Fred Hutchinson Cancer Research Center. Ever since oxygen was discovered in the early 1770s, “scientists have recognized it as essential to life,” he says. What the 18th-century scientists didn’t know is that oxygen is essential to life in a surprisingly nonbinary way. “Yes, if you take away oxygen, you can kill the animal,” Roth says. “But if you further reduce the oxygen, the animal is alive again, but it’s suspended.”

He has shown that this works in soil nematodes, which are alive in air with as little as 0.5 percent oxygen and are dead if you reduce the oxygen to 0.1 percent. But if you then proceed quickly to a much lower level of oxygen—0.001 percent or even less—the worms enter a state of suspension where they need significantly less oxygen to survive. It’s their way of preserving themselves during extreme deprivation, a bit like animals hibernating in winter. These oxygen-starved, suspended organisms appear to be dead but not permanently so, like a gas cooktop with only the pilot light on.

Roth is trying to get to this pilot-light state by infusing experimental animals with an “elemental reducing agent,” such as iodide, that greatly decreases their oxygen needs. Soon he’ll try it in humans too. The goal is to minimize the damage that can occur from treatments after heart attacks. If iodide slows oxygen metabolism, the thinking is, it might help avoid the blowout injury that sometimes comes with treatments like balloon angioplasty. At this lower setting the damaged heart can just sip the oxygen coming in through the repaired vessel, rather than get flooded by it.

Life and death are all about motion, according to Roth: In biology the less something moves, the longer it tends to live. Seeds and spores can have life spans of hundreds of thousands of years—in other words, they’re practically immortal. Roth imagines a day when using an agent such as iodide, a technique that will soon be studied in early clinical trials in Australia, can give people that immortality “for a moment”—the moment they most need it, when their heart is in serious trouble.

Such an approach would not have helped Pérez, whose heart never stopped beating. The day after her devastating CT scan, her obstetrician, Somer-Shely, tried to explain to Pérez’s stunned and frightened parents, Berta and Modesto Jimenez, that their beautiful daughter—the lively young woman with sparkly eyes who adored her little girl, had a passel of friends, and loved to dance—was brain-dead.

There was a language barrier. The Jimenezes’ first language is Spanish, and everything the doctor said had to be filtered through a translator. But the real barrier wasn’t language. It was the concept of brain death itself. The term dates to the late 1960s, when two medical developments coincided: high-tech, life-sustaining machinery, which blurred the border between life and death, and organ transplantation, which made clarifying that border especially urgent. No longer could death be defined in the traditional way, as cessation of breath and heartbeat, since ventilators could provide both indefinitely. Is a patient on a ventilator dead or alive? If you remove the ventilator, when can you ethically retrieve the organs to transplant into someone else? If a transplanted heart starts beating again in a new chest, was the heart donor really dead in the first place?

To address such thorny questions, a Harvard panel met in 1968 to define death in two ways: the traditional way, by cardiopulmonary criteria, and a new way, by neurological ones. The neurological criteria, which are now used to determine “brain death,” involved three cardinal benchmarks: coma or unresponsiveness, apnea or the inability to breathe without a ventilator, and the absence of brain-stem reflexes, measured by bedside exams such as flushing the ears with cold water to see if the eyes move, poking the nail bed to see if the face grimaces, or swabbing the throat and suctioning the bronchia to try to stimulate a cough.

It’s all quite straightforward, yet also counterintuitive. “Brain-dead patients do not appear dead,” wrote James Bernat, a neurologist at Dartmouth’s medical school in New Hampshire, in the American Journal of Bioethicsin 2014. “It is contrary to experience to call a patient dead who continues to have heartbeat, circulation, and visceral organ functioning.” His article, meant to clarify and defend the concept of brain death, appeared just as two controversial patients were making headlines: Jahi McMath, a California teenager whose parents refused to accept the diagnosis after the girl experienced a catastrophic loss of oxygen during a tonsillectomy, and Marlise Muñoz, a brain-dead pregnant woman whose case differed from Pérez’s in a significant way. Muñoz’s family didn’t want anything done to sustain her body, but hospital staff overruled them, because they thought Texas law required them to keep the fetus alive. (A judge eventually ruled against the hospital.)