A tree is a very common datastructure. I stumbled upon this subject because of a requirement that is needed for a mathematical solution. This solution needs to keep track of changes that start with a single change but can have multiple effects. Each effect, i.e. change, can result in turn into multiple changes, hence the need of a tree structure.
I have been looking for a best practice to create a major single page application using the excellent Feliz library. I also heavily borrowed ideas from a book by the author of Feliz, Zaid Ajaj. The application in mind is intended to be the Dutch National Pediatric Emergency app used for acute interventions in pediatric medical emergencies. The main purpose of the application is to provide all calculations necessary based on age and weight of the patient.
Software development for Informedica has been a one man show. Maybe that’s about to change. However, this means that the development process has to change as well. In this blog I propose a development process to enable this change. The fork and pull model will be used to cooperate.
Het ergste wat een ouder kan overkomen is dat zijn/haar kind kritisch levensbedreigend ziek wordt. De Kinder IC is dan vaak de laatste strohalm. Gelukkig wordt de behandeling steeds beter op de IC, maar daarmee worden ook de uitdagingen steeds groter. Een zeer belangrijke uitdaging daarbij is groeiende complexiteit van zorg die geleverd moet worden aan toenemend complexe patiënten.
To Err is Human. This is how a landmark paper from 2000 starts, recognizing that “the problem is not bad people in health care–it is that good people are working in bad systems that need to be made safer“*. The consequences of errors can be described by adverse events. Adverse events that are related to medication and or drug/fluid incidents constitute about 20% of all types of adverse events, which makes this the second most common type of adverse event†.
At the PICU (Pediatric Intensive Care Unit) of the University Medical Center Utrecht, we have been working the last years to make the data from our PDMS (Patient Data Management System) available for research. This has resulted in a multilayered system with generic possibilities to extract data and transform those data into a convenient flat table format (as described in a previous post).
Machine Learning (ML) (or Artificial Intelligence AI) is trending. Publications with regard to ML are on the increase in medical literature.
Specifically critical care medicine generates huge amounts of detailed data. In a recent article steps are described to enable use of ML in daily clinical practice. This blog will describes an actual working implementation of the first step in ML to be used in clinical practice.
Indentation is important in F#, as it defines the code blocks and the separate code elements. Using Feliz to define a view, in a Fable.React application, the indentation matters just as much, and even more.
Fantomas is a beautiful tool that automatically formats your code and ensures consistency, also in indentation. But using code like with the Feliz library, there some serious drawbacks.
In a previous post, a setup using an F# script to perform machine learning with Microsoft.ML is described. A very import aspect for a successful model is picking the right features that will predict the label, i.e. the exposure that is associated with outcome. In this post a simple F# feature selection algorithm is described that automatically figures out which features result in the ‘best’ model.
Machine learning is the new kid on the block and has become accessible with the advent of the Microsoft.ML library. In medicine machine learning hasn’t been used that much. Most scientific epidemiological research still relies on established statistical analysis. The core principle, however, is the same. Known exposure is used to predict outcome. In Pediatric Intensive Care, prediction of mortality is used to benchmark and monitor performance. In the Netherlands, for this purpose, data is gathered at a national basis. This blog will discuss a machine learning setup to analyze data using a data set with 13.793 PICU admissions.