Innovative Segmentation Strategies for Melanoma Skin Cancer Detection

Innovative Segmentation Strategies for Melanoma Skin Cancer Detection

Bibliographic Details
Main Author: Munnangi, Anirudh
Language:English
Published: University of Cincinnati / OhioLINK 2017
Subjects:
Biomedical Research
Skin Cancer
Melanoma
Neural Netoworks
Self Organized Feature Maps
Mixture of Experts
Image Segmentation
Online Access:http://rave.ohiolink.edu/etdc/view?acc_num=ucin1510916097483278
id ndltd-OhioLink-oai-etd.ohiolink.edu-ucin1510916097483278
record_format oai_dc
spelling ndltd-OhioLink-oai-etd.ohiolink.edu-ucin15109160974832782021-08-03T07:04:34Z Innovative Segmentation Strategies for Melanoma Skin Cancer Detection Munnangi, Anirudh Biomedical Research Skin Cancer Melanoma Neural Netoworks Self Organized Feature Maps Mixture of Experts Image Segmentation The purpose of this project is to research innovative segmentation algorithms that will be the part of skin cancer detection process. As a part of the thesis, two application specific modeled algorithms have been designed to perform the process of segmentation, which is the second step in the overall process of classification of the image into various cancerous categories. A novel attempt to use a clustering based algorithm to address a segmentation task has been attempted and achieved through this research. Images have been considered in the gray scale mode and an attempt has been made to extract maximum results without color information. Both algorithms developed involve training and testing phases. Also, they are inspired by the power of Neural Networks. Once the segmentation is done, various performance metrics have been calculated and reported along with visual aid regarding how well the segmentation occurred. The performance has also been compared with the commonly used methods in image segmentation and the advantages as well as performance factors are well critiqued and documented to provide a holistic view related to the usage of such algorithms in the concerned topic of skin cancer segmentation. Experimental testing has also been done with images having pre-known ground truth information and the resulting segmented portions as well as quality has been shown. 2017 English text University of Cincinnati / OhioLINK http://rave.ohiolink.edu/etdc/view?acc_num=ucin1510916097483278 http://rave.ohiolink.edu/etdc/view?acc_num=ucin1510916097483278 unrestricted This thesis or dissertation is protected by copyright: some rights reserved. It is licensed for use under a Creative Commons license. Specific terms and permissions are available from this document's record in the OhioLINK ETD Center.
collection NDLTD
language English
sources NDLTD
topic Biomedical Research
Skin Cancer
Melanoma
Neural Netoworks
Self Organized Feature Maps
Mixture of Experts
Image Segmentation
spellingShingle Biomedical Research
Skin Cancer
Melanoma
Neural Netoworks
Self Organized Feature Maps
Mixture of Experts
Image Segmentation
Munnangi, Anirudh
Innovative Segmentation Strategies for Melanoma Skin Cancer Detection
author Munnangi, Anirudh
author_facet Munnangi, Anirudh
author_sort Munnangi, Anirudh
title Innovative Segmentation Strategies for Melanoma Skin Cancer Detection
title_short Innovative Segmentation Strategies for Melanoma Skin Cancer Detection
title_full Innovative Segmentation Strategies for Melanoma Skin Cancer Detection
title_fullStr Innovative Segmentation Strategies for Melanoma Skin Cancer Detection
title_full_unstemmed Innovative Segmentation Strategies for Melanoma Skin Cancer Detection
title_sort innovative segmentation strategies for melanoma skin cancer detection
publisher University of Cincinnati / OhioLINK
publishDate 2017
url http://rave.ohiolink.edu/etdc/view?acc_num=ucin1510916097483278
work_keys_str_mv AT munnangianirudh innovativesegmentationstrategiesformelanomaskincancerdetection
_version_ 1719453201391943680

Similar Items