Essay On 3d Mammography

Abstract— Radiologists have only the 2D mammographic reports to detect the abnormalities present, which do not give the feel that how the abnormalities exactly look like. Hence there is a need of representing the abnormalities (like mass, tumor, micro calcification) in 3D to increase the accuracy in detection. A literature survey is done about the various techniques used in segmentation and 3D representation of the abnormalities.

Keywords—mammography, 2D, 3D representation, segmentation. I. INTRODUCTION

Mammography is a breast imaging technique that detects any abnormalities present. It is usually taken in two views viz.,
Cranio-caudal(CC) view that is taken by projecting the rays from ‘head to toe’ and Medio lateral

This work uses the mammographic images taken in CC, MLO and lateral views.
It allows the specialist to detect the tumors. The specialist can draw on the mammogram, by using the mouse, to segment out the suspicious region. The lines drawn are get adjusted to the contour of the suspicious region. Incorporating the expert’s role in the detection of the tumor helps in enhancing the accuracy of detection. The segmented tumor and the contour of the breast are mapped in two orthogonal planes. They are so adjusted that the tumors from the two views coincide. The contours of the breast are then used for reconstructing it in 3D with triangular mesh. The tumor is represented in 3D by using a bounded volume box of shapes ellipsoid, sphere. The location where it presents is marked in the 3d triangular mesh breast model.

V. Ramalingam et al [9] proposed a method of segmenting the masses using only the single view image i.e. either CC or MLO view. Initially, the image is preprocessed to remove the artifacts, pectoral muscle and noise. The breast contour is segmented by means of binarization. It involves dividing the image into two parts: background and