LUNG NODULES SEGMENTATION IN CHEST CT BY LEVEL SETS APPROACH

Archana A, Amutha S

Abstract: Segmenting Lung nodules from a chest CT image have vital interest for medical applications like diagnosis and surgical planning. Here the focus is to segment the small lung nodules from lung CT scan by level set approach that uses signed distance function. A general lung nodule shape model is proposed using implicit spaces as a signed distance function which is fused with the image intensity statistical information. Mapping of shape model to image domain is done by a global transformation (inhomogeneous scales, rotation, and translation) and is matched with the image implicit representations to handle the alignment process. Shape alignment process is handled by evolving transformation parameters through gradient descent optimization that marks the boundaries of the nodule. Image intensity as well as prior shape information is used overlay images to segment out the nodule. A nonparametric density estimation approach is employed to handle the Statistical intensity and background region. This algorithm aims to segment the nodule irrespective of its type or location.

Keywords: SEGMENTATION

DOI: https://doi.org/10.15623/ijret.2015.0426006