Intelligent decision support in home health care

Important results from the project

The aim of the project was to develop intelligent decision support that enables efficient home monitoring of patients with chronic diseases. Time series with data from twenty years of research projects in mainly COPD and heart failure have been the basis for model building with the help of machine learning in order to identify factors that allow for prediction and early action to prevent deterioration of the disease.

Expected long term effects

The central idea has been to use AI and machine learning to use data and scientific evidence from previous studies with Phoniro´s digital Health Diary to support individualized remote monitoring of symptoms and status parameters as well as monitoring of intake of emergency medication. The studies of decision parameters for COPD and heart failure with the help of machine learning have been carried out as planned and are reported openly in scientific publications. However, it has not been possible to implement the results as part of a product within the frame of the project.

Approach and implementation

Studies were performed with machine learning algorithms to understand the predictive potential of our baseline data. We also created statistically sound synthetic data that was generated on baseline measurements and used this synthetic data to create more robust ML models. In the project, we have also collected new data with self-monitoring of patients requiring care with COPD. Heart failure data and modeling for palliative home healthcare have also been addressed.