Pattern Recognition

The interesting thing about dermatology is that it is all about visual pattern recognition. A dermatologist just glances at a rash and knows what it is. Looking at the image above, you immediately know it as psoriasis or have no idea. Even though there is a systematic approach to rashes, nearly always, a dermatologist makes a diagnosis without thinking through it systematically. Though I could tell you about the shiny silvery scale, the well-demarcated border, the location on an extensor surface, a picture is a thousand words, and the sight makes the diagnosis. Sometimes they can't even explain exactly how they know something is a mole rather than a skin cancer; it's a culmination of experience. "I've just seen so many, I know this needs to be biopsied," they say.

Nevertheless, this thought process is not unique to dermatology. Although it is most easily distinguished because it is visual pattern recognition, all of medicine involves this heuristic thought process. If you say "headache, fever, photophobia," my first instinct is lumber puncture to rule out meningitis. But if you instead said, "headache, nausea, photophobia," I'm thinking more migraine (but certainly meninigtis is a concern). Lightheadedness, chest pain, and tachypnea in a 50 year old man concerns me for a heart attack (myocardial infarction), but the same triad in a 20 year old woman triggers anxiety in my head. This fascinates me; how do we learn these shortcuts, how do we put together these stories, how do we come to recognize these patterns of signs and symptoms to reach a conclusion without working systematically? And, how good is that way of thinking?

Image of psoriasis shown above is from Wikipedia, shown under GNU Free Documentation License.