Premature brain damage image classification algorithm based on association rules

摘要

Aiming at the particularity of MRI images, the ARC algorithm which is suitable for medical image processing is present. The new method introduces the bi-support association rules based on FP-tree. In the process of generating the rule of association class and constructing the FP-tree, the maximum support is introduced on the premise of minimal support. It can make the support of discovered rules be between the maximum support and minimal support. so through the bi-support, the interference in the rules is excluded. Experiments shows that this ARC method could achieve better results in practice than CMAR in the operational efficiency and classification accuracy. It is very effective in CWMD digital image classification.