The Woman Who Walked Away

I have been a radiologist for twenty-two years. I have read approximately 300,000 imaging studies. I have testified as an expert witness in medical malpractice cases. I teach residents how to interpret MRIs, CTs, and plain films. I am, by training and temperament, a diagnostician. I trust what the images show because I have seen them be right tens of thousands of times.

Maria came to our emergency department after a car accident. She was thirty-one, otherwise healthy, and had been the driver in a high-speed collision on the interstate. Her initial trauma workup was remarkable only for what was found on MRI: a complete transection of the spinal cord at T6, with retraction of the cord ends visible on sagittal imaging. In plain English: her spinal cord had been severed completely. The two ends of the cord were separated by approximately three centimeters on the scan. There was no ambiguity. I reviewed the images myself, as did two of my colleagues. All three of us agreed on the finding.

A complete spinal cord transection at the thoracic level means permanent paralysis from the chest down. There is no recovery. There is no surgical repair. The prognosis is categorical: the patient will never walk again, will have no sensation below the level of injury, and will require lifelong assistance with bowel and bladder function. These are not probabilities — they are certainties. Spinal cord axons do not regenerate in humans. Every textbook, every journal article, every clinical experience confirms this.

Maria was told her prognosis. Her family was called. A consult was placed with the spinal cord rehabilitation service. The plan was to stabilize her medically and transfer her to inpatient rehab within the week.

Three days after admission, Maria told her nurse that she could feel her toes. The nurse documented it in the chart, noting that the patient claimed to have regained sensation in her feet. I reviewed the note and dismissed it as denial — a common psychological response to catastrophic injury. Patients often report phantom sensations or insist they can feel movement that isn't there. I'd seen it before.

On day five, she moved her left toes. I saw the video on the nurse's phone — unmistakable, voluntary movement of the toes on command. I ordered a repeat MRI, fully expecting to find that the initial reading had been wrong — that the transection was incomplete, that there was a thin bridge of cord tissue we had missed, that the diagnosis had been too aggressive.

The repeat MRI showed an intact spinal cord at T6. Not a partially healed cord — an intact cord. There was no evidence that a transection had ever occurred. I compared the two scans side by side on my workstation. The initial scan clearly showed a three-centimeter gap between the cord ends. The follow-up scan showed a continuous, normal-appearing cord with no defect, no scarring, and no evidence of surgical repair — because none had been performed.

Maria walked out of the hospital on day twenty-three. She used a walker for the first week, then transitioned to a cane, then walked unaided. I have presented this case at four radiology conferences. I have shown the images to hundreds of radiologists, neurologists, and neurosurgeons. I have never received a satisfactory explanation. I do not expect to.