The Shadow in the MRI

I noticed the shadow in March. It appeared on the twelfth axial slice of an abdominal MRI I was reading — a small, roughly spherical density in the left upper quadrant that didn't correspond to any anatomical structure. I dismissed it as an artifact. MRI artifacts are common: motion artifacts from patient breathing, susceptibility artifacts from metal implants, truncation artifacts from the reconstruction algorithm. Every radiologist learns to recognize and disregard them.

But the shadow appeared on the next scan I read. And the next. And the next. By the end of March, I had documented the same artifact on seventeen scans — all abdominal MRIs, all from the same machine, all on the same twelfth slice.

I called the MRI technologist. She ran the standard quality control phantom — a cylindrical test object filled with a solution that produces a predictable signal pattern. The phantom scan was clean. No artifact. She recalibrated the machine. She ran a second phantom. Clean again. She told me the machine was fine.

The shadow continued to appear. I pulled the images from all seventeen scans and overlaid them. The shadow was identical in every case — same size, same density, same location. If it were a calibration error, it would vary. If it were a patient-related artifact, it would vary. But seventeen different patients, scanned on seventeen different days, produced exactly the same artifact on exactly the same slice? That was not an artifact. That was a signal that something was generating a consistent response in the magnetic field at exactly the point where the twelfth slice passed through the bore of the magnet.

I walked down to the MRI suite and stood in the bore of the magnet — the long, narrow tube where patients lie during scans. I ran my hand along the interior wall, feeling for anything that might be causing the artifact. At the position corresponding to the twelfth slice, I found it: a small, almost imperceptible indentation in the magnet housing. About the size of a fist. As if something had struck the interior wall with considerable force and left a permanent impression.

The machine had been installed six years earlier. According to hospital records, it had never been damaged, never required repair, never shown any abnormality on monthly QA. I showed the indentation to the chief of radiology, who showed it to the manufacturer's service engineer, who said he had never seen anything like it in thirty years of servicing MRI machines. They replaced the housing panel. The shadow disappeared.

I never learned what caused the indentation. The manufacturer's report noted "unexplained physical deformation of magnet bore housing" and recommended replacement. The cost was covered under warranty. Nobody was hurt. But I have a recurring dream in which I am the patient, lying in the bore, and the twelfth slice passes through my chest, and I feel something — a pressure, a warmth, a presence — that I cannot explain. And then I wake up.