The story behind 'coarse-grain': A case for scaling of observations in Pathology
It was a day like any other day when I thought of listening to a podcast at Player FM. Being a philosophy aficionado I searched for some content that would be entertaining and educational in the domain of philosophy, so I listened to a conversation between a physicist and a behavioural neuroscientist. The show was called Sean Carroll's Mindspace and the episode, 67 Kate Jeffery On Entropy, Complexity, and Evolution. This episode, to my delight, had a ring of biological philosophy to it. The speakers went on about the topic of their research and the physicist and the neuroscientist went on a spree of discussing their research questions with each other which in another world would've never been discussed. To tell you something about the guest, Kate Jeffery received her Ph.D. in behavioural neuroscience from the University of Edinburgh. She is currently a professor in the Department of Behavioural Neuroscience at University College, London. She is the founder and Director of the Institute of Behavioural Neuroscience at UCL. As I listened to the speakers speak I thought I heard the host utter, "coarse-grain". He spoke in detail about entropy in the physical world and how computationally supple the various events are made by coarse-graining. The term caught my fancy and I researched its meaning and applications in the field of physics wherein computations based in atomic and molecular scale, if complex, are made a bit simpler thanks to coarse-graining. The term essentially means handling computational complexity in a bit simpler and more thicker terms. The word thicker means, if one is coarse-graining at the Quantum scale, one computes the state atomically, if at the atomic or molecular scale, one computes at a state wherein all molecules that share similar properties are clubbed together and denoted as groups. This progression provided me with an insight about scales at which observations can be made. Being a Pathologist, I thought of imbibing it in our observations. Voila! There I had it. The entire scale unfolded as an insight into the nature of scales at which observations in Pathology are made. The scales at which our branch is observing diseases tells something about how the reductionist impulse in science is taking precedence over and above the emergent.
Reductionist impulse lets a scientist observe a concept at smaller and smaller scales. Emergent impulse lets a scientist observe concepts at bigger and bigger scales. There is an inherent contradiction in these two observational modes. The addition of smaller units doesn't simply make the whole. The whole is much bigger than the sum of its parts. That being said, I devised a progression of scales wherein mutually exclusive observations can be made and noted in a format that depicts clearly what observational work has been performed at various scales. Thus, I posited a progressive scale in an ascending order, as follows:
To tell you the truth, much has been sorted in the way CAP protocols suggests to write observations, but none can be used as a working model, more or less like a list of observations made during the whole process of testing, in a clean and sorted style that provides with detail, as a checklist, all the phenomena observed during reporting of samples. CAP essentially is meant for grading and staging as an arbitrary format for reporting for various tissue samples. It doesn't categorically state the scale at which an observation has been made and hence is lackadaisical as far as scales and anomalous observations are concerned. The addition of scales will open up avenues of researching into the unpredictable manifestations diseases make when they are being observed real-time. One can then successfully add a qualitative aspect to the research in Pathology, which is quite rare. Phenomena like autolysis, processing, staining, reporting, and various correlations can be done if one has a handy tool. The scaling of observations can be that handy tool which can serve as an audit of sorts for the in-house team that handles the samples in quality check. Thus, we can have a statistical analysis at which scaling of observations can be useful, not to forget how easy and free of verbosity the reports will be once the residents get a knack of reporting the observations using scales.