IMPROVED LUNG NODULE CLASSIFICATION USING MULTI-CLASS ARTIFICAL NEURAL NETWORK WITH BACK PROPAGATION ALGORITHM
Abstract
vioft2nntf2t|tblJournal|Abstract_paper|0xf4ff38572b000000048f050001000100
This paper designed a novel method to overcome the problem of the lung nodule overlapping adjacent structures. We developed a lobe segmentation algorithm for identifying lung lobes CT images. To find reliable method for nodule detection is an important problem in medicine. It requires efficient automatic method to perform segmentation and detection. The identification of tumor region involves extraction of lobar fissures from the input CT images which makes use of two phases. In the first phase the fracture region is identified. In the second phase the found fissure are extracted. There is some nodule-like object in testing data detected by algorithm and not included in ground truth information. These are probably nodules missed by human. We designed a novel method to overcome the problem of the lung nodule overlapping adjacent structures. The result Obtained show that the proposed work can help the surgeons to identify the lobar fissures correctly to locate the lung region before they plan for the surgery. It reduces the computation time and complexity. Our system was developed with Faculty Hospital, Motol and Prague and in future should be used there. In order to improve the performance of the proposed approaches some future enhancements could be necessary in the present research work.

Authors
S Vimalnath1, G Ravi2
Paavai Engineering College, India1, Sona College of Technology, India2

Keywords
CT image, Lung Nodule, Classification, ANN
Yearly Full Views
JanuaryFebruaryMarchAprilMayJuneJulyAugustSeptemberOctoberNovemberDecember
000000000000
Published By :
ICTACT
Published In :
ICTACT Journal on Image and Video Processing
( Volume: 10 , Issue: 2 , Pages: 2083-2088 )
Date of Publication :
November 2019
Page Views :
153
Full Text Views :

Creative Commons License
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.