A Game of Shadows 1


by Edison McDaniels, MD

A truism: the evolution of neurosurgery has largely paralleled the evolution of imaging of the nervous system. Great strides have been made since the 1970’s, when the first CT scans became available and it thus became possible to localize an anatomic lesion within the central nervous system with great accuracy.

There was imaging available before the 1970’s, of course. The period before about 1974 or ’75 might be called the era of shadows—for that is what the imaging amounted to. In fact, even today, plain x-ray imaging is really a game of shadows.

Different tissues of the human body are composed of different densities, with the most dense being the teeth, followed by the bones, and the least dense being the air-filled structures, especially the lungs. The denser the tissue, the more prominent the shadow on x-ray.

With the exception of the teeth, bones cast the most prominent shadow on x-ray. Indeed, the shadow of a bone may be so dense as to obscure everything both in front and behind it (remember, plain x-ray is a 2D affair—it’s as if the body has been flattened to a single plane, like a photo, which is exactly what an x-ray is; as an example, if you take a picture of a person, the person obscures everything behind them—in the case of an x-ray, the bone obscures everything behind and in front of it).

2015-05-25 11.30.43

This is an xray of the lumbar spine. There are actually 4 screws, but because they overlap each other it looks like only two.

Reading a plain x-ray is thus, largely, about learning pattern recognition—what sorts of shadows various tissues cast. Sometimes, the shadows overlap, the densities summing to appear to show something when in fact there is nothing there. This is known as a superfluous confluence of shadows.

The situation is made all the more complicated by the idea that the shadows are often not representative of the structures themselves, but only of their relative presence or absence. What? One example will do nicely to illustrate the situation.

Consider the state of the art in neuroimaging prior to the mid-1970’s, pneumoencephalography (the title image is a pneumoencephalogram). This test, invented in 1919 by one of the founding fathers of modern neurosurgery, Walter Dandy, involved replacing the CSF with air injected into the ventricular system within the brain. The ventricles are normally fluid filled and thus, on plain x-ray at least, have a density similar to that of the surrounding brain (for the purposes of plain x-ray, brain and water have nearly the same density). As such, a normal or even abnormal ventricle is not visible on a plain x-ray and certainly is not discernable from the surrounding brain.

Dandy, for various reasons, had the idea to replace the CSF of the ventricles with something either more or less dense than the surrounding brain and thus make them visible on x-ray. Why? Because of the whole shadow idea. If the ventricles could be made visible, their shape would be discernable. If their shape was discernable, the prying eyes of the physician would be able to see if they were enlarged, say by hydrocephalus. Or, more to the point, if they were distorted—that is the ventricles were shifted out of their normal position or shape—by the presence of a tumor. In visualizing this, the location of the tumor would thus be suggested and the tumor localized not directly by the shadow of the tumor but indirectly by the abnormal shadow of the ventricle.

And localizing the tumor was of course, the holy grail of neurosurgery since one could not safely or even comfortably operate upon a tumor in the human brain if one did not have a fairly good idea of where it was to be found in the first place.

Edison McDaniels, MD, is a board certified neurosurgeon practicing in the American South. Follow him on twitter @surgeonwriter and read his fiction on Amazon in both paperback and kindle. 

Click here for Part 2: More on pneumoencephalography, and the development cerebral angiography—one of the great advances in medicine. 

Send to Kindle