Research areas
The main objective of the CISIAD is to do research on probabilistic graphical models (PGMs), such as Bayesian networks and influence diagrams, and their applications in diagnosis and decision making, especially in medicine. The CISIAD is also interested in the application of statistical multivariate analysis and economical evaluation techniques
Research fields
- Artificial intelligence and expert systems
- Probability
- Decision theory
- Applied statistics
- Medical diagnosis and decision analysis
- Economic evaluation
- Health technology assessment
Basic research
- Construction of PGMs from objective data (databases, epidemiologic studies, etc.) and human expert knowledge
- Algorithms for the evaluation of PGMs (computation of probabilities and decision making)
- Development of software tools for the construction and evaluation of PGMs, in accordance with the principles of software engineering
- Design of interfaces for real-world applications: graphical user interfaces (GUIs), database interfaces, communication interfaces (Internet, LANs...), etc.
Aplications
Medical applications:- Diagnosis of heart diseases by echocardiography
- Prognosis and decision making in palliative care units
- Diagnosis of prostate cancer and other urologic diseases
- Diagnosis of rhinopharyngeal cancer
- Diagnosis of liver diseases
- Decision making in pneumology (lung cancer)
- Decision making on cataract surgery
- Diagnosis of failures in fossil power plants
- Computers security