Loss, Risk and Posterior Distributions
For a while, I’ve been thinking about the deployment of predictive algorithms in clinical decision support. Specifically, about the difference between what we understand about a model’s performance from the publication describing it and how this might be less informative when deployed. In short: what is the value of knowing that a model has good balanced accuracy or a high area under the ROC curve when sat with a patient and using the tool to make a clinical decision.
https://danwjoyce.netlify.com/post/what-we-do-in-the-ehr/
Clinical State and Dynamical Systems
In this series of blogposts, we look at some models of clinical state. The motivation is to document exploratory work with a colleague (Nick Meyer, who runs the SleepSight study) as we try and apply some theoretical ideas – for example (Nelson et al. 2017; Scheffer et al. 2009) – to ‘real-life’ data.
Similarity Maps for Antipsychotic Medications - Part Four
Some thoughts and experiments on how to best represent mult-dimensional data on receptor binding in one comprehensive map.
Similarity Maps for Antipsychotic Medications - Dopamine and Serotonin Receptor Affinities : Part Three
Visualising affinity profiles and similarities between anti-psychotic medications.
Similarity Maps - Visualizing Antipsychotic Affinities : Part Two
Explore the dopamine receptor affinity / binding profiles for common antipsychotic medications using a geometric concept of similarity.
Visualizing Antipsychotic Receptor Affinity : Part One
Different medications bind to receptors with different affinities … This post explains visualizing this property for common medications in the antipsychotic class.
Thumbnail Credit : https://commons.wikimedia.org/wiki/User:Pleiotrope
Reductionism gets bad press and by my reckoning, is too often used to discredit something that might deserve more careful attention had it not been summarily dismissed as reductionist by it's opponents. Here's my opinion on why ...