AI-DPA

The main objective of the project is to analyze medical treatment pathways for prostate cancer using process mining. In simple terms, the aim is to reconstruct typical “treatment pathways” from a large number of individual medical records. The project is based on a real, large registry dataset from Rhineland-Palatinate that covers cases across multiple institutions and represents the entire end-to-end patient journey.

The research project AI-DPA is a joint initiative of Hochschule Mainz – University of Applied Sciences and the University of Koblenz. The project focuses on the use of Artificial Intelligence (AI) and Machine Learning to analyse and interpret unstructured data and processes in two- and three-dimensional application scenarios.

The project brings together research from two main areas: computer vision and 3D data analysis, as well as Predictive Process Monitoring.

One focus is on the analysis of 3D point clouds for indoor and outdoor environments. Point clouds are 3D representations of physical spaces that can be captured using sensors such as LiDAR scanners or smartphones. AI-based methods are used to identify and interpret objects and spatial structures within these data. This research can support applications such as indoor navigation, smart buildings, facility management and accessibility planning. For example, the analysis of a 3D-scanned building can help identify stairs, doors, obstacles and other elements that are relevant for creating more accessible environments.

The second focus is Predictive Process Monitoring. Here, Machine Learning is used to analyse processes, recognise patterns and predict what may happen next. Applications include chatbots, user click behaviour and processes in media organisations. The aim is to better understand user behaviour and processes and to support their optimisation.

A key aspect of the project is the combination of Deep Learning and knowledge-based methods. Technologies such as Neural Radiance Fields (NeRF) and self-learning ontologies are investigated to improve the semantic interpretation of complex and incomplete data.

Overall, AI-DPA aims to contribute to the digital transformation of different sectors, including media, healthcare, urban development and accessibility. By turning complex data into meaningful information and predictions, the project seeks to provide a foundation for more efficient processes and informed, data-driven decision-making.

Explore more!