An efficient association rule-based method for diagnosing ultrasound kidney images

The objective of this work is to develop and implement a computer-aided decision support system for an automated diagnosis and classification of ultrasound kidney images. This approach combines automatically extracted low-level features from images with high-level knowledge given by a specialist in order to suggest a diagnosis of a new kidney image. The proposed method distinguishes three kidney categories namely normal, medical renal diseases and cortical cyst. The preprocessing technique applied on the images eliminates the inconsistent data from the US kidney images. Then feature extraction process is applied to extract the features from the US kidney images. Feature selection and discretization process is done on the extracted features that reduce the mining complexity. The proposed method uses a new algorithm ARCKi is a new associative classifier. This classifies the given image to suggest a diagnosis with high values of accuracy. The performance of our approach is compared with multilayer back propagation network in terms of classifier efficiency with sensibility, specificity and accuracy.