Jump to content

Dr. Marcela Saeb's Blog

  • entries
  • comments
  • views

Learning and thinking through the pattern of images Dermatopathology



To start this blog I will quote Dr. Paola Domizio [i]“ What makes a good pathologist? First of all, I think you have to be good at pattern recognition... you have to be thorough in your thinking...[/i]”

So what is pattern recognition, how can someone create a visual thinking, how is memory involved in this process? As dermatopathologists we go through this process every time we look into the microscope or digital images.

How can I identify a pattern or recognize it? A pattern is the collection or sum of certain stimuli or features that should be the same as what was sensed (visualize) before. There has to be previous information learned through books where patterns or features of images are shown and match written concepts, this gives the necessary tools to identify diseases. These features are process in the cortex of the right side hemisphere and creates a visual memory, and through repetition turns into visual thinking. It is said that visual thinking is essential for certain activities, such as driving, flying, catching a ball, calculating speed trajectory time, to give some examples.

Do pathologists turn out to be visual thinkers and visual learners? It seems that this is exactly what we do. We think and learn through the features or patterns of images. We learn those patterns and we go through slides recognizing them, which together with the clinical information, supports the diagnosis of a disease.

Whenever there is no recognition of any pattern, even reviewing books or recent literature, we might be in front of a new entity as Dr. Juan Rosai has said: ‘If you see something that's doesn't fit any disease you know or that is in the books, don't assume automatically that you don't recognize it because you are no good. Maybe it is something that has not yet been described’."

Visual memory and visual thinking are essential in the process of recognizing patterns, however, they do not work alone, there is a scientific method that has organized and classified patterns of pathology images in different clinical settings, which makes this process not just a technique but an art of science.


Recommended Comments

Dr. Phillip McKee


Marcela, I like your blog as it does summarise what we do. If one has not encountered a lesion before, there is no way that one can recognise it. After all, if someone hasn't pointed to a cat and told you that it is a cat you will never know that a cat is a cat. Similarly if it has only 3 legs is it still a cat? This implies that not only do you learn by example and by good instruction by good teachers/mentors but you also have to gain personal experience and learn to recognise that the many variations on the theme might not change the diagnosis. For example, a lichenoid keratosis often has parakeratosis with plasma cells and an occasional eosinophil in the dermal infiltrate. If these features are absent it doesn't mean that the lesion isn't a lichenoid keratosis. The clincial information of a solitary lesion makes the diagnosis.
Link to comment
Dear Phillip, thank you very much for your comment. What most amazed me about this, is that we acquire a "repertoire" of images so well organized, that as soon as we see "the cat" even with one leg or not "whiskers" at all, it is still "a cat".
Link to comment
  • Create New...