Kelly Thomas woke up in a Florida hospital four years ago with no recollection of the car accident that had robbed her of the ability to walk.

Thomas, an active college student who had barrel raced in rodeos, moved to Kentucky for a year to try out a research study that she hoped would retrain her spinal cord on how to walk.

In February, a trainer who had been scooting along the floor at Thomas's feet, helping her move her legs and place her feet as she used a walker, stopped and stood up.

"What are you doing?" Thomas asked, alarmed.

"You're doing it," the trainer, Rebekah Morton, told her. "You don't need me."

Thomas hesitated, and then took a step on her own. Then another. She froze.

"I'm like, 'That just happened.' I've been working so hard for four years. I got emotional," Thomas recalled.

Thomas, now 23, is one of several people with spinal cord injuries who are standing, taking steps and - in her case - walking without assistance, thanks to an experimental combination therapy.

In a research study at the University of Louisville, Thomas and three others had a device surgically implanted on their spinal cords to stimulate electrical activity, accompanied by months of daily physical therapy.

In the New England Journal of Medicine, researchers from the University of Louisville reported that two of the subjects could stand and take limited steps at the end of the study, and two were able to walk independently - Thomas and another patient, Jeff Marquis.

The research was supported by a charitable foundation, the University of Louisville Hospital and device maker Medtronic.

A simultaneous case report published in Nature Medicine reported that a single patient with a spinal cord injury at the Mayo Clinic was also able to take steps and walk with trainer assistance with electrical stimulation and intensive physical therapy.

"The history of spinal cord injury research is we have 50 years or more of essentially failed trials, with no positive findings," said David Darrow, a neurosurgery resident at the University of Minnesota Medical School who was not involved in either study but has also been implanting a stimulator device in people with spinal cord injuries.

"This is sort of a new era."

Darrow said there are caveats to the newest studies, and a slew of unanswered scientific and medical questions.

This is still a tiny number of patients, with variable injuries, so it is impossible to know how well the intervention will work in the broader population of people with spinal cord injuries.

There are also plenty of questions about how the technique works, which will be untangled only as a broader community of researchers begins to study the approach in more patients.

But the results are a powerful proof of concept, built on a surprising case report by the Louisville research team in 2011.

Researchers took a medical device developed to manage chronic pain and implanted it to stimulate the spinal cord of a paralyzed patient. After rehabilitation sessions, that patient learned to stand and regained some voluntary control of his leg movements.

Susan Harkema, associate scientific director of the Kentucky Spinal Cord Injury Research Center at the University of Louisville who pioneered the technique in people, said that the devices are implanted well below the site of the injury.

This isn't a case of patients regrowing some severed connection in the spine. Instead, Harkema sees these studies as heralding a gradual shift in how experts think of the spinal cord - as able to learn new ways to walk, with the right combination of training and electrical stimulation.

"The basis of this work is that the spinal circuitry is sophisticated and really has the same properties that the brain does in many ways, and in the context of this study, really what is shown is it has the capability of relearning to walk in the right conditions," Harkema said.

The intervention is not like flipping a switch. First, study subjects were given about two months of intense physical therapy and training, to make sure that alone wouldn't restore function.

After they had the device implanted, they began a rigorous course of daily therapy, as a team of therapists began to train their bodies and minds on how to step again.

Thomas said that it was far from intuitive at first, as she would get cues like "toe up" or "shift your weight" or "pull your knee up."

She began stepping with her right leg, on a treadmill, on her third session after the implant. Getting the left leg took longer.

"It was extremely, extremely hard at first," Thomas said.

"I couldn't talk to anybody, couldn't look at anybody - I was completely focused on my body. Now, I can walk and talk, and it's not as much of a struggle. It's still not easy, and it's not completely natural."

Thomas has been able to integrate her new abilities into her daily life after returning home to Florida. She puts her walker into her car, brings the remote control that allows her to operate the stimulator, and goes out on her own to the library, to restaurants and to get her nails done.

The researchers at the Mayo Clinic saw similar results with their technique; a patient with a complete spinal cord injury was able to take steps and walk with trainer assistance.

Kelly Thomas learning to walk again (Tom Fougerousse/University of Louisville)

In both studies, the patients needed the stimulator to be on to walk, helping to rule out the idea that this was spontaneous recovery. Taking steps could be done only when people were trying to move their legs.

"The important point is this technology may be able to give back functional control, to stand and take independent steps. So it really gives hope to people who are faced with paralysis," said Kendall Lee, a neurosurgeon at the Mayo Clinic.

The hope is that as these techniques begin to be tested in more places and in more patients, the demand will drive the technology to be improved and tailored for the intervention.

The researchers readily admit they are not engineers and hope to see a stimulator developed with this application in mind, instead of finding new purposes for an off-the-shelf device.

The technique will also need to be tried in more diverse people; the study subjects were in their 20s and early 30s, and it had been about two or three years since their injuries.

Many people with spinal cord injuries are older, and more years have passed since they became paralyzed.

Darrow has implanted the stimulator in someone whose injury happened 17 years ago, and the oldest subject in his ongoing study is in their late 50s.

Darrow is interested in whether the electrical stimulation has effects on its own, without the rehabilitation - on voluntary movement, and on other health issues, such as low or uncontrolled blood pressure.

"I saw their work and thought this was really cool. There's just tremendous potential," Darrow said.

"If we could get more effort in the area and bring the people with the skill sets to really change this field ... you can make a lot of progress."