Head of the Centre: Ágnes Vathy-Fogarassy, Associate Professor
Email: vathy.agnes@mik.uni-pannon.hu
The most important medical data sources (e.g., hospital information systems, GP information systems, registers, national financial databases) store a huge amount data about medical treatments of the whole population. However, the information contained in these databases is often unexploited, and their use is usually limited to the retrospective and prospective collection of data, to the utilization in the operational patient care processes, and production of basic statistical analysis.
The goal of our research and development center is to perform healthcare analyses, develop optimized solutions and healthcare information systems, which are based on the exploitation of the healthcare data assets. To explore the knowledge in the data, we use the most dynamically developing discipline of information technology, namely the data science and the toolkit provided by the operations research.
During our researches, we develop a number of healthcare-specific methods and algorithms that implement effective information exploration and optimization adapted to healthcare. The applied data mining, network analysis, process mining, statistics, machine learning, deep learning algorithms and related visualization techniques can effectively help doctors and researchers discover relationships and patterns that are specific to patients, health care or diseases, and to prove or reject hypotheses. Furthermore, the healthcare sector offers optimization opportunities at many points when tested by formal means. Thus, by examining healthcare processes with the help of mathematically precise integer and mixed-integer models, optimal scenarios and bottlenecks of the system can be identified, providing improving steps.
The continuous cooperation of the healthcare specialists (physician, research physician) and IT specialist (data scientist) is an important basis for our activity. During our researches, we work closely together with the specialists of different research directions, who play an important role in defining the directions of researches, validating results and designating application areas.
R&D activities:
- data mining analysis of healthcare data
- development of domain-specific data-driven machine learning methods for leveraging healthcare data assets
- IT support for personalized medical care
- development of intelligent medical decision support systems
Main applications and research areas:
- Performing basic statistical analyses (calculation of population-level statistical indicators, incidence analyses, analysis of waiting and sojourn times, etc.)
- Retrospective exploration of healthcare processes:
- exploration of typical treatment patterns
- identification of treatment deviations
- certification of protocol tracking
- Optimization of healthcare treatment processes:
- identification of bottlenecks of medical care
- proposal of optimized treatment processes
- adapting graph theory framework to optimization methodologies in healthcare data under uncertainties
Development of optimized healthcare decision support systems - Complication assessment:
- incidence analyses of complication assessment
- identification of risk factors
- development of scoring systems
- Survival analysis
- Correlation studies
- Efficiency studies
- Selection of control groups for retrospective case studies
Key R&D Accomplishments:
- Identification of the main risk factors for heart failure complications in patients undergoing anthracycline treatment and development of a validated scoring system based on these risk factors.
- Identification of drugs used in the prevention of cardiotoxicity.
- Development of a cardiology decision support system.
- Comparative analysis of cancer survival trends in Hungary during the periods 2001–2005 and 2011–2015 based on the population-based cancer registry.
- Classification of chemotherapy treatment effectiveness in lung cancer patients based on N-glycan data.
- Development of a multi-level process mining methodology for the analysis of healthcare service events.
- Development of exact mathematical and artificial intelligence-based scheduling solutions for outpatient care scheduling.
- Development of a text mining methodology for the automatic structuring of echocardiography reports.
Projects:
- Collaborative Hospital Information Platform: creating a medical system to support process-oriented inpatient care (2021-2024)
- Development of intelligent cardiological process-based decision support system (2016 September – 2021 January)
- Development of intelligent and inclusive health information and decision support framework (2014 – 2017)
Partners:
- Asseco Central Europe Magyarország Zrt.
- National Healthcare Services Centre
- GE Healthcare
- Ferenc Csolnoky Hospital
- State Hospital for Cardiology
- Translational Glycomics Research Group (University of Pannonia)