Please use this identifier to cite or link to this item:
Title: Brain Tumor Detection using Association Allotment Hierarchical Clustering
Researcher: Tamilmani G
Guide(s): Sivakumari S
Keywords: Engineering and Technology
Computer Science
Computer Science Interdisciplinary Applications
University: Avinashilingam Institute for Home Science and Higher Education for Women
Completed Date: 2021
Abstract: In medical domain, the Computer Aided Diagnosis (CAD) system is utilized in diagnosing many newlinediseases like brain tumor, breast cancer, etc. But, initial stage of brain tumor detection is difficult newlineand accurate diagnosis is very challenging. If brain tumor is medically suspicious, a radiological newlineassessment is needed to find its location, size and the impact on surrounded region. These issues newlinecan be addressed by data mining techniques. The proposed automatic diagnosing method involves newlinecertain processing steps: pre-processing, image-segmentation, feature extraction, feature newlineoptimization and process of tumor classification. The Mutual Piecewise Linear Transformation newlinemethod has been proposed by combining Contrast Stretching (CS) and Linear Stretching (LS) newlinetechniques for pre-processing of brain tumor images. The principal task of pre-processing is to newlineenhance the quality of the image and improves image parameters such as eliminating unnecessary newlinenoise, optimizing the signal-to-noise-ratio (SNR) with visual appearance of image, maintaining its newlineedges, softening the internal portion of the region. newlineThe image segmentation is the process of segmenting the image as numerous regions newlinedepending on its qualities like color, size, texture and roughness. Association Allotment Hierarchical newline(AAH) Clustering algorithm is proposed for segmentation. It is used to discover the patterns of newlineperiodic image. Feature extraction is to extract the quantitative information like shape, texture, color newlineand contrast. The different textures and statistical features are extracted from the segmented output newlineafter the process of image segmentation. The statistical measures such as mean, variance, range, newlineStandard Deviation (SD), entropy, skewness and kurtosis are computed for the extracted features. newlineThe Grey Wolf Optimization (GWO) approach has been proposed for feature selection process. newlineFurther, the proposed Neuro-fuzzy classifier is employed to classify the brain tumor as normal or newlineabnormal. newlineThe performance of the proposed methods has been evaluated and a
Pagination: 130 p.
Appears in Departments:Department of Computer Science and Engineering

Files in This Item:
File Description SizeFormat 
01_title.pdfAttached File64.43 kBAdobe PDFView/Open
02_certificate.pdf152.23 kBAdobe PDFView/Open
03_acknowledgement.pdf11.9 kBAdobe PDFView/Open
04_table of contents.pdf49.68 kBAdobe PDFView/Open
05_list of tables, figures and abbrevatiions.pdf80.15 kBAdobe PDFView/Open
06_chapter 1.pdf349.21 kBAdobe PDFView/Open
07_chapter 2.pdf142.74 kBAdobe PDFView/Open
08_chapter 3.pdf462.57 kBAdobe PDFView/Open
09_chapter 4.pdf422.96 kBAdobe PDFView/Open
10_chapter 5.pdf788.27 kBAdobe PDFView/Open
11_chapter 6.pdf450.36 kBAdobe PDFView/Open
12_chapter 7.pdf10.86 kBAdobe PDFView/Open
13_references.pdf131.39 kBAdobe PDFView/Open
80_recommendation.pdf247.42 kBAdobe PDFView/Open
Show full item record

Items in Shodhganga are licensed under Creative Commons Licence Attribution-NonCommercial-ShareAlike 4.0 International (CC BY-NC-SA 4.0).

Altmetric Badge: