Bimodal system for interactive indexing and retrieval of pathology images

Abstract

The prototype of a system to assist the physicians in differential diagnosis of lymphoproliferative disorders of blood cells from digitized specimens is presented. The user selects the region of interest (ROI) in the image which is then analyzed with a fast, robust color segmentel: Queries in a database of validated cases can be formulated in terms of shape (similarity invariant Fourier descriptors), texture (multiresolution simultaneous autoregressive model), color (L*u*v* space), and area, derived from the delineated ROI. The uncertainty of the segmentation process (obtained through a numerical method) determines the accuracy of shape description (number of Fourier harmonics). Tenfold cross-validated classification over a database of 261 color 640 x 480 images was implemented to assess the system performance. The ground truth was obtained through immunophenotjping by flow cytomety. To provide a natural man-machine interface, most input commands are bimodal: either using the mouse or by voice. A speech synthesizerprovides feedback to the user: All the employed computational modules are context independent and thus the same system can be used in a large variety of application domains.

Content based MRSI image indexing and retrieval involving a fuzzy descriptor

“…Korn et al proposed a system for the fast and effective retrieval of tumor shapes in mammogram X-rays [5]. Comaniciu et al described a system that aims to help physicians in the diagnosis of lymphoproliferative disorders of blood [6]. Glatard et al introduced a Texture Based Medical Image Indexing and Retrieval for Cardiac Images [7].…”

CT Image Retrieval Using Tree Structured Cosine Modulated Wavelet Transform

“…Korn et al, [6] proposed a system for fast and effective retrieval of tumor shapes in mammogram X-rays. A system which aims at helping physicians in the diagnosis of lymphoproliferative disorders of blood is described [7]. Nevertheless, a description of the clinical use of such systems is very rare, except for the systems ASSERT [9], which is dedicated to HRCT images of the lung and includes information from physicians-such as anatomical landmarks and pathology bearing regions-and IRMA which proposes a multi-step approach for the [13] author has proposed a standard wavelet based method for CT image retrieval.…”

Content-based cell pathology image retrieval by combining different features

“…In another system reported in [10], the images in the database consist of a single tumor in the center without any background texture. The system presented in [11] aims at helping physicians in the diagnosis of lymphoproliferative disorders of the blood. Shape and texture features were used to characterize the regions of interest delineated by the user.…”