Updated: Jan 28, 2019
Series: Lessons from optical illusions
When an image has ambiguous interpretations.
What do you see?
Do you see a vase? .............................................then look at the edges on either side and you might change your perception.
Do you see two faces? ...................... think of the edges of each face forming the outline of a symmetrical vase to change your perception.
This mental 'flipping' of perspective produces what is called a 'Gestault switch' between seeing the vase and seeing the opposing faces.
This 'Gestault switch' can happen in pathology too.
This is a biopsy of a skin rash from a 70 year old man's arm.
If you are not used to looking at skin biopsies under the microscope, I can tell you that this is a very abnormal biopsy.
For comparison, this is what normal skin looks like:
Step one: there is something wrong with this biopsy. The next step is to find out what is wrong. But it's hard to tell exactly what that is because there is so much inflammation. In fact there is so much inflammation that it obscures the skin surface. The demarcation looks almost blurry, with the usual distinctive demarcation not readily seen.
You might not be able to guess at where the skin surface falls, so I've added the demarcation, crudely drawn.
One way to find out what is wrong is by determining where the problem comes from:
A problem from the outside, growing in (the skin surface growing down)?
A problem from the inside, growing out (cells within the dermis growing out)?
My colleague saw this case first and was concerned about a cancer of the epidermis - a problem growing from the outside in.
My colleague showed it to me to get my opinion, and instead I thought of a problem coming from the inside, with inflammatory cells being drawn to the skin, like a lymphoma.
Who was right?
We don't have to only look at these images and guess.
Now that we generated a list of possible diagnoses (greatly simplified here as an inside or outside problem), we need to gather more information. Our next step involves using tools that demonstrate the origin of the problem. These are called immunohistochemical stains. Broadly speaking this test labels specific cells so that we can tell what type of cell it is. You can think of these as a molecular paintbrush, highlighting cells with pigments, usually brown, that reveal the markers on the cells and therefore the true nature of the cells.
My colleague ordered stains that highlight only the cells of the epidermis, the keratinocytes (the outside).
That's what is seen here in brown - epidermal keratinocytes. They are all staying behind the junction, and they are NOT leaving the skin surface.
Different stains were used to prove or disprove my hypothesis about the problem - that it is an inside-out problem.
This next picture shows a stain for white cells (lymphocytes), specifically large and malignant white cells (CD30).
This shows these large and irregular cells crossing over into the top skin surface. It is like a dermatitis, where lymphocytes cross over into the skin. But these cells are abnormal and malignant - they are invading into the epidermis. This is a cutaneous lymphoma. And when it is full of CD30 positive cells, we will call it a Large Cell Lymphoma.
The stains labeled the cells and demonstrated where the problem arose - this was an inside-out kind of question. It was, after all, a vase and not two faces.
What information we chose to gather was based upon knowing how our initial perception of the case could be wrong.
We generated a differential diagnosis - a set of possible diagnoses that make sense for the patient and the case. In this case, the differential I shared with you (an inside or outside problem) was a simplified version of the actual differential. After we have this list of possibilities, we then test out these diagnoses - sometimes with stains, others with molecular tests.
Imagine -- what if when I look at a case, I only wanted to prove my understanding of it, and not seek to determine what is it's true nature? This would be like only trying to verify it is a vase, rather than seeking to develop tests that demonstrate the true nature of the image. Proving the rightness of one explanation is a hallmark of pseudoscientific approaches to health. Instead, the practice of pathology is rooted in the scientific tradition, where we seek to disprove our own hypotheses.
The lesson from many optical illusions is that we must take in information to offset visual and mental distortions. However imperfectly we do it, if we don't seek to question and demonstrate our information, we will be ruled by our initial perceptions. And our brains, wired as they are for greatness, have traps of perception we must strive to overcome.