For Project Enquiry +91 9791 044 044

EXISTING SYSTEM

PROPOSED SYSTEM

EXISTING CONCEPT:

·         An edge-based LSM based on the distance regularized level set evolution (DRLSE) and the LRLSM with the automated selection of localization radius (LRLSM_A) were single-image-based and the COGRLSM based on the foreground similarity and background consistency.

·         The COLRLSM with constant weights of localized energies from image pairs (COLRLSM_C) was based on our algorithm.

PROPOSED CONCEPT:

          Segmentation of pheochromocytomas in Contrast-Enhanced Computed Tomography (CECT) images is an ill-posed problem due to the presence of weak boundaries, intratumoral degeneration, and nearby structures and clutter.

          The initialization method is simple and feasible and highly reduces the manual labor for physicians.

EXISTING  TECHNIQUE :

·         CONSTANT METHOD

PROPOSED ALGORITHM:

          ADAPTIVE METHOD

TECHNIQUE DEFINITION:

·         The constant method was not flexible to handle both weak and distinct boundaries.

·         The existing algorithm didn’t gain much more satisfied results which attributed to the adaptive use of the globalized information of image pairs.

·         It is difficult to segment tumor in images with complex foreground and background.

ALGORITHM DEFINITION:

         The adaptive method was more flexible to handle both weak and distinct boundaries.

          The proposed algorithm gained much more satisfied results which attributed to the adaptive use of the localized information of image pairs.

          It is easy to segment tumor in images with complex foreground and background.

DRAWBACKS:

·         The co-segmentation method with constant localized energies was not flexible and robust enough and had less impressive segmentations.

·         COGRLSM analyzed global image information that was more robust against poor initialization that local methods. However, it was prone to boundary leakage in the cases with weak boundaries.

·         COGRLSM produced unfavorable performances in terms of accuracy and robustness.

ADVANTAGES:

          Automated or semi-automated segmentation methods would have improvements in efficiency and accuracy.

          Fast, readily available, highest spatial resolution.

          Very high sensitivity in detecting primary adrenal Pheochromocytomas, equivalent to MRI.