Background: The diagnosis of diseases like Attention Deficit Hyperactive Disorder and cervical cancer needs an efficient segmentation technique to identify the regions of problematic areas with low cost and CPU time.
Method: The Proposed RD - ACM methodused to identify the ADHD and cervical cancer affected areas. In this method, the acquired input images are enhanced, converted into binary images using optimal threshold value andthe connected components are extracted using label matrix. The initial contours are generated and taken as the zero level set.Neumann Boundary is generated to specify the size of the image where the Level Set Evolution (LSE) process will take place. Then the image is smoothen by Heaviside and Dirac delta function. The proposed method provides a piecewise constant solution is derived by the introduction of diffusion term into LSE.A Two-step splitting method iteratively solves the RD-LSE equation is introduced to first iterate the LSE equation, and then solve the diffusion equation and to regularize the level set function obtained in the first step. The process is repeated until the final contours of the objects are extracted.
Results: Significant promising segmentation results are noticed for all type of images with boundary antileakage. It can be applied to solve both variational as well as partial differential equation based level set methods. The application of Reaction- Diffusion term ensures stability, and thus the complex and costly reinitialization procedure is completely eliminated from LSE.
Conclusion: Anovel medical image segmentation technique using optimal threshold Reaction-Diffusion Active Contour model (RD - ACM) is found to be morfe promising technique in the diagnosis of ADHD.
Keywords
