Automatic Analysis of Patient Notifications with Intelligent Models

Based on the grouping of patient complaints, new knowledge can be extracted. Based on this new knowledge, appropriate actions can then be identified which contribute to improvements in the quality of care.

Aim of the project

This project aims to conduct a first study with the goal of categorizing and analyzing the information content of patient reports and from there extract knowledge about abuses in healthcare. Today, the content of the reports is subjective descriptions of what the patient experienced and the information is largely unstructured.

Implementation

We plan to use self-learning (unsupervised) clustering algorithms to group the patient reports into groups (so-called clusters) based on their information content. By evaluating different types of clustering algorithms, we will investigate to what extent it is possible to group the notifications based on the common problems they describe and the patients' own suggestions for appropriate actions.