Abstract—Due to the large volume of data generated in
healthcare organizations, the use of data mining techniques
becomes essential for improving the quality of care, physician
practices and disease management. However, expert knowledge
is not based only on rules, but also on a mixture of knowledge
and experiences.
It is in this context that we set the involvement of data mining
techniques and CBR to support medical decision making in order
to optimize the time and benefit from the experience of experts.
We propose a support system for medical decision-making based
on CBR and data mining. This system allows, from a database of
examples, engaging a method of Symbolic induction and Cellular
Inference Engine (MIC) for the construction of a case retrieval
model.
To evaluate this new approach we have customized the platform
jCOLIBRI with a real case base about the treatment of tuberculosis.