Publication

Abstract : Content Based Image Retrieval (CBIR) is the application of computer vision techniques to retrieve the most visually similar images from the image database for any given query image. The visual characteristics of a disease carry diagnostic information and oftentimes visually similar images correspond to the same disease category. In this paper we aim at building an efficient Content Based Echo Image Retrieval (CBEIR) system for 2D Echo (2DE) and Color Doppler Flow (CDF) image modalities. From 2DE images, features such as dimensions of cardiac chambers (area, volume, ejection fraction, etc) are extracted; whereas texture properties, kurtosis, skewness, edge gradient, color histogram, etc., are extracted from CDF images. Hence, this forms a multi-feature descriptor which then is used to retrieve similar images from the database. Some of the major contributions of our work are: modified K-Means segmentation algorithm coupled with PL/SQL and External Procedures to achieve speed, accurate detection of cardiac chambers using active contour model, efficient method to extract color segment from CDF images, and flexible multifeature model. These domain specific low-level features are very important to build a reliable and scalable CBIR model. The feature database is a set of quantitative and qualitative features of the images. Our image database is populated with diverse set of approximately 623 images of normal and abnormal patients acquired from a local cardiology Hospital. Exhaustive experimentation has been conducted with various input query images and combinations of features to compute the retrieval efficiency which are validated by domain experts. It has been shown through Recall-Precision graphs that the proposed method outperforms compared to others reported in the past.