inst/digilogger_gui/concept.md

In michbur/digilogger: Data Mining and Machine Learning Tool for POCT devices

During the analysis using POCTs, data is generated continuously. This is particularly the case with multiparametric systems. With multiparametric systems such as VideoScan technology, the following values, among others, are recorded simultaneously in one sample:

• DNA and protein biomarkers
• Time of measurement
• Age
• Gender.

In principle, the DNA and protein biomarkers can already be validated and clinically proven or are still at an experimental stage. The latter are of particular interest because they can be assumed to be suitable for diagnostic applications in the future.

On the basis of the measured values, classes (e.g., "normal", "borderline", "diseased") can in principle be determined. This knowledge can be used for further purposes.

The aim of the digilogger technology is the continuous acquisition of measurement data and the collection of measurement data in a database which is used for machine learning. Models are continuously developed and compared to previous knowledge. For example, threshold values for new (experimental) biomarkers are to be determined automatically. In addition, the software will support the classification of patients.

michbur/digilogger documentation built on Nov. 4, 2019, 6:51 p.m.