Conrad was 17 months old when Dave, his grandfather, was babysitting him at their home in Temple, Texas. The two had been playing in the pool and went inside for a break. Dave set to unloading dishes in the dishwasher, unaware that Conrad had snuck back outside.

As he finished the dishes, Dave looked out the window and noticed something odd. There was what looked like a floating bundle of clothes in the swimming pool. It was his grandson.

Fortunately, Conrad responded to cardiopulmonary resuscitation (CPR), but it’s unclear how long his lungs—and his brain—went without oxygen.

Drowning is the second most common cause of accidental death in children to age four. As in Conrad’s case, CPR is fortunately very successful, with 66 percent of nearly drowned children surviving. But even when resuscitated, the seconds and minutes that the brain is deprived of oxygen come at a great cost. This type of damage is known as anoxic brain injury.

Anoxic brain injury is a clinical term that indicates damage to the brain that occurs due to lack of oxygen. There is a spectrum of injury ranging from complete recovery to minor to widespread brain damage. Within this spectrum lies what is known as the disorders of consciousness, with the extent of damage being proportional to the loss of consciousness.

In the case of nearly drowned children, the injury is frequently thought to be widespread. Nearly drowned children are labeled “minimally conscious” or even in a “persistent vegetative state” (with no consciousness) and the prevailing medical prognosis is grim: treatment and recovery is difficult if not impossible.

Conrad was left with profound brain damage. He remained comatose, unresponsive for weeks. Following his emergence from the coma, parents and doctors struggled to understand what part—if any—of Conrad’s brain function remained.

Liz Tullis, Conrad’s mother, explained to UT Health San Antonio, “When Conrad survived his accident, I was not given much hope or guidance; in fact I was encouraged to institutionalize Conrad. Other families were encouraged to withdraw care.”

One complication was that it was difficult to tease out what normal, “adult” brain functions had been lost because of the injury and which simply hadn’t “grown out” yet—recall that Conrad was less than two years old when his accident happened. Even before the accident we wouldn’t have expected him to hopscotch or spell.

Conrad’s doctors felt his brain was severely damaged and that, if he improved at all, he might only be able to take food by mouth or do simple movements like roll over or hold his head up. Whether Conrad would be able to think and reason wasn’t on the table.

That Conrad’s ability to think was uncertain reflects a clinical malaise surrounding the detection and prediction of internal mental states. It also reflects a very specific clinical limitation: the clinical interview requires movement.

Whether listening to your heart beat (contracting, relaxing, contracting, relaxing), observing you breathe in and out, tracking the way your eyes focus, and perhaps most importantly, evaluating your mental and emotional state based on the words you speak, the clinician’s skill rests on some form of movement. Subtract movement and the clinician can gather no data.

But what if a patient can’t move? Or at least move in a meaningful way? Does lack of movement necessarily imply lack of thought? (Non moveo ergo non cogito ergo non sum?)

With time, Conrad regained a limited ability to move—not coordinated, purposeful movements like hopscotch, but rather the ability to wiggle and squirm. He couldn’t speak. To all outward clinical evaluations, he had reached his “optimum treatment potential.”

When their insurance would no longer cover Conrad’s physical therapy and treatments, the Tullis family could fortunately self-pay and persist with the hope that their baby would continue to improve.

Liz had a sense that Conrad was “in there” and wanted to sort out whether there was evidence for that. She approached Dr. Peter Fox, a neurologist and neuroscientist at the UT Health San Antonio’s Research Imaging Institute, to see if there was something he could do to help Conrad, perhaps some avant-garde brain-computer interface that could improve his state.

Dr. Fox said he could offer no treatment, but with the help of modern brain imaging methods, he could see what brain systems Conrad had intact. He offered to design a study of nearly-drowned children with anoxic brain injury. With the help of his then-graduate student, Dr. Janessa Manning, a pediatric neuroscientist currently at the N.I.H.’s Perinatology Research Branch, they began to peer inside Conrad’s brain and inside the brains of 9 other children with near-fatal drowning. They chose to use resting-state functional Magnetic Resonance Imaging (resting-state fMRI).

Resting-state fMRI was an ideal method to study these children’s brain function because it is task-free, meaning that each child only had to remain still in the MRI scanner. The method captures the brain’s intrinsic activity in the absence of a specific task. Resting-state fMRI has shown that brain networks that subserve motor and even cognitive functions like language, memory and emotion are continuously and dynamically active in the resting brain.

The results of their study, published July 31, 2017 in the journal Human Brain Mapping, suggest there is great hope for children with anoxic brain injury. The pattern of brain damage wasn’t as widespread as previously thought and indicates much higher levels of brain function.

In ten out of ten of the children studied, damage was most severe in important motor networks deep inside the brain. This explained the absence of purposeful movement in children like Conrad.

Also in all ten of the children, brain networks involved in language, emotion, and memory were seemingly preserved. This means that those children were awake and alert even though their absence of coordinated movement labeled them “minimally conscious.” Liz was right, Conrad was “in there.”

Dr. Mariam Ishaque, an MD-PhD student (and neurosurgery applicant) who spearheaded the project, was surprised how well Liz’s clinical impression was borne out by the imaging data. She recalled to me that “one child couldn’t talk, or control his movements. But he made eye contact with us. I could tell that he was annoyed that we were keeping him awake [they scanned the children at night]. You could tell he knew what was going on. It was striking.”

It appeared that near-fatal drowning causes an anoxic brain injury that is more akin to focal stroke than to widespread brain damage. These children had a clinical condition more akin to locked-in syndrome (where some patients have enough consciousness to write a book!) than a persistent vegetative state, where the brain function is flat-lined.

“This is a new syndrome,” Dr. Fox remarked to me. “It’s not taught in medical school…and this study presents a new level of sophistication for functional MRI. We have developed a method of using resting-state fMRI to diagnose individual patients. We hope that these results will help bring resting-state fMRI and per-patient analyses into the clinical arena.”

What this exciting study promises isn’t a new treatment, but a way to direct these children to different therapies. Soon after a near drowning, this study suggests that children could benefit from neuroprotective agents or even similar therapies that stroke victims receive. In addition, it suggests that in some cases, parents and doctors are justified in continuing support.

Nearly drowned children like Conrad could very well be “fully-conscious,” able to think and reason, able to understand social situations and humor and Dr. Seuss.

As Dr. Manning explained to me, “If these children are in fact locked-in, then they are part of a much larger population who share similar types of brain damage, like children with cerebral palsy. This presents opportunities for clinicians and educators to develop methods to intellectually challenge these children who are very much conscious, just in a different way.”