From Data to Knowledge in the Oncology Domain (DKOD)

Data mining methods can be used for extracting specific medical knowledge to improve the quality and efficiency of patient care. Mining retrospective data stored in registers is a new challenge in the medical informatics research domain.

Data pre-processing is an essential step in preparing the raw data for knowledge extraction. Data cleaning, transformation and data reduction are important parts of this process. There are many different data mining methods and each one has its own strengths and drawbacks. We need a fast, accurate and explainable mining method with efficiency comparable to that of domain experts. Computer-interpretable guidelines (CIG) can support clinicians at the time of decision-making by presenting suggestions and plans for both diagnosis and treatment according to existing clinical guidelines.

This research project focuses on studies with the goal of developing a decision support system in the oncology domain. When oncologists encounter new patients, the system can be used to support them in their decision-making by presenting both extracted knowledge achieved from the mining as well as advice from existing guidelines in the field. This study is supported by grant No. F2003-513 from FORSS, The Health Research Council in the South-East of Sweden.

Research Group

Nosrat Shahsavar, PhD	Project manager, Supervisor
Hans Åhlfeldt, PhD	Co-supervisor
Amir R. Razavi, MD	PhD Student
Hans Gill, MSc	Research Engineer
Olle Stål, PhD	Co-supervisor
South-East Swedish Breast Cancer Study Group	Domain Experts
List of related publications