The rapid increase in digital imagery, including satellite, medical, and general -purpose images, has highlighted the need for efficient image analysis and classification systems. Image Mining plays a crucial role in extracting meaningful insights from vat volumes of image data. Association Rules (AR) mining, traditionally used for pattern recognition in transactional databases, has emerged as a promising approach for uncovering complex relationships within image datasets. This paper provides a comprehensive review of the application of (AR) in image classification, with a particular focus on the medical domain. It compares traditional and state-of-the art techniques, highlighting their strengths, limitations, and the challenges associated with image mining using AR. Furthermore, the paper explores future research opportunities, such as integrating deep learning with association rule mining, to advance the field of image classification.