Search or add a thesis

Advanced Search (Beta)
Home > Multi-Resolution Tranaform Based Feature Extraction Techniques for Differentiating Glioma Grades Using Mri Images

Multi-Resolution Tranaform Based Feature Extraction Techniques for Differentiating Glioma Grades Using Mri Images

Thesis Info

Access Option

External Link

Author

Zia, Razia

Program

PhD

Institute

National University of Sciences & Technology

City

Islamabad

Province

Islamabad

Country

Pakistan

Thesis Completing Year

2018

Thesis Completion Status

Completed

Subject

Electrical Engineering

Language

English

Link

http://prr.hec.gov.pk/jspui/bitstream/123456789/10763/1/Razia_Zia-Elec_Engineering_NUST_2019_21.03.2019.pdf

Added

2021-02-17 19:49:13

Modified

2024-03-24 20:25:49

ARI ID

1676727794906

Asian Research Index Whatsapp Chanel
Asian Research Index Whatsapp Chanel

Join our Whatsapp Channel to get regular updates.

Similar


Medical image processing is one of the most attention gaining research areas that utilizes the technology for improving the quality of human life through a more precise and rapid diagnosis systems. This thesis focuses on computer assisted diagnosis of brain neoplasms which is amongst the most fatal cancers. Though, their exact cause is still unknown but early detection anddiagnosisofcorrectneoplasmtypeisveryimportantforpatient’slifeandfurthertreatment planning. Currently, the treatment of brain neoplasm depends on clinically observed symptoms, appearance of radiological tests, and often the microscopic examination of neoplasm’s tissues (histopathology or biopsy report). Magnetic Resonance Imaging (MRI) is the state of art technique to diagnose brain neoplasms and monitor their treatment. It provides a noninvasivewaytoimprovethequalityofthepatient’slifethroughamoreaccurateandfastdiagnosis and with minor side-effects, leading to an effective overall treatment. However, MRI does not provide any information about exact type and grade of neoplasm. The final decision is based on biopsy report of patient which is considered as gold standard, despite all risks associated with surgery to obtain a biopsy. With rapid advancement in technology, the researchers are continuously working on computerized techniques or computer assisted diagnostic tools to provide fast identification, correct diagnosis and effective treatment of brain neoplasm. The aimofthepresentthesisistodesign,implement,andevaluateasoftwareclassificationsystem fordiscriminatingthreegradesofbrainneoplasmonMRI.Limitedbrainneoplasmimagedata isoneofthebiggestissuesinthisresearchareabecausecollectionofthistypeofdatarequires years and years. Normally, we find studies working on images of some specific hospital orwebsite. Inaddition,directcomparisonofthesestudiesisnotpossiblebecauseeachstudyhad worked on different types of neoplasm and various sizes of image data. We have addressed this issue by proposing a new image cropping technique for handling images of different dimension for the same classifier. This new system is capable of handling image datasets from different institutions with various image sizes and resolutions for comparing, regulating and sharing of research. It is also observed, that lesser training and testing images in a particular class of neoplasm badly effect the classification accuracy. By using this generalized system, moreimagesamplesofaneoplasmclasscanbetakenfromotherinstitutionsorwebsitestoimprovetheclassificationaccuracy. ForclassificationofMRIimages,majorityoftheresearchers haveworkedonstatisticalfeaturesofneoplasmregionbutmulti-resolutiontransformsforfeature extraction, are not much explored. Besides this, classification of normal and pathological brain is mostly addressed but very few studies are found on multi-classification of different neoplasm types. The main objective of this thesis is to explore the performance of different multi-resolution transform based feature extraction techniques for multi-classification problem of brain neoplasm type (grade II, grade III and grade IV gliomas). Discrete Wavelet Transform (DWT) is one of the most popular multi resolution transform, extensively used as feature extraction technique for binary (normal vs abnormal brains) brain neoplasm classification systems. In this thesis, a stationary and time invariant Non Subsampled Contourlet Transform (NSCT) with Gray Level Co-occurrence Matrix (GLCM) is used for computation of feature vector in brain neoplasm classification system. This NSCT-GLCM based classification system is also compared with conventional DWT-GLCM based classification system, for the same experimental setup. It is found that NSCT-GLCM based system perform better than DWT-GLCM based system. For further improvement in neoplasm discrimination accuracy, in last algorithm, a multi resolution transform based hybrid feature extraction technique is introduced. This hybrid technique is comprised of conventional DWT, NSCT and GLCM. The quantitative performance analysis showed that hybrid feature extraction technique per formed much better than the previous two techniques (DWT-GLCM and NSCT-GLCM) with the highest accuracy of 88.88%. The developed brain neoplasm classification techniques can better assist the physician’s ability to classify and analyze pathologies leading for a more reliable diagnosis and treatment of disease.
Loading...
Loading...

Similar Books

Loading...

Similar Chapters

Loading...

Similar News

Loading...

Similar Articles

Loading...

Similar Article Headings

Loading...

مولانا ابوالاعلیٰ مودودی

مولانا ابوالاعلیٰ مودودی
۲۲؍ ستمبر ۱۹۷۹؁ء کو مولانا ابوالاعلیٰ مودودی کی وفات سے یہ برصغیر ہی نہیں بلکہ پوری اسلامی دنیا بھی ایک ممتاز متکلم، ایک مضطرب مفکر، ایک دیدہ ور مفسر، ایک بے چین شارح دین، ایک قابل قدر ترجمان شریعت اور ایک بلند پایہ مصنف سے محروم ہوگئی۔
انھوں نے ۷۷ برس کی عمر پائی، ان گنت کتابوں کے مصنف ہوئے شروع میں ان کی تحریریں نکلیں تو جدید تعلیم یافتہ طبقہ کو محسوس ہونے لگا کہ ان میں وہ ساری باتیں ہیں جن کی تلاش ان کے ذہن کو ہے، جب وہ مغربی افکار کے ابلیس کی تلبیس سے دب کر اپنے تذبذب اور تشکیک کی بناء پر اسلام کو جامد اور غیر متحرک پارہا تھا تو اس کو مولانا مودودی کی تحریروں کے ذریعہ سے یہ احساس ہوا کہ کوئی اس کے تاریک اور منجمد ذہن پر ہلہ بول کر اس کو جھنجھوڑ رہا ہے، مولانا مودودی کی تحریروں میں شگفتگی بھی ہوتی، دلنشینی بھی، حرارت ایمانی کی گرمی بھی، نظر و فکر کی گرم جوشی بھی، منطقی دلائل کے یقین کی پختگی بھی، عالمانہ انداز میں کلام پاک اور حدیث کی جاندار تعبیر بھی، اسلام کے ناقدوں اور خصوصاً فرنگی معترضوں کے خلاف جارحانہ حملے بھی اور یورپی طرز فکر کی جرأت مندانہ تحدی بھی، اس لئے وہ شوق سے پڑھی گئیں، ان کا علم ایک بحرخار تھا، جس سے انھوں نے بقول استاذی المحترم مولانا سید سلیمان ندوی، ’’یورپ کے ملحدانہ افکار‘‘ کے خلاف ایک بند باندھنے کی کوشش کی اس میں ان کے قلم کی بے مثال قوت ان کا پورا ساتھ دیتی رہی۔
رفتہ رفتہ وہ ایک خاص مکتب فکر اور تحریک کے بانی قرار پائے جس سے صالح قیادت کے ذریعہ وہ مسلمانوں کی معاشرتی اور سیاسی زندگی میں طاقتور دینی روح دینی اخلاص اور دینی قدریں پیدا...

فلاح معاشرہ ميں عائلى زندگى كا كردار اسوہ حسنہ کی روشنی میں

Islam gives complete guidance in every walk of life and has described clear rules in this regard. Family life is a core institution of society and playes a vital role in the welfare and prosperity of society. Prophet Muhammad (SAW) presented His comprehensive and beautiful character regarding family life. In this paper an attempt has been made to discuss the corelation between family life and society and how does family life plays its role in the development of socity

Exploring the Contribution of Mentoring in Improving Teaching and Learning

Producing highly skilled, competent teachers in developing countries, such as Pakistan and Kyrgyzstan, requires an on-going teacher education, training and support. A number of research studies indicate that mentoring provides a highly useful service for teachers in their continuing education. As such, mentoring has been widely accepted as a process to support beginning teachers' professional development. However, little is known in developing countries about what mentoring is, and to what extent it brings improvement in teaching and learning. Therefore, in this particular case study was designed to explore the contribution of mentoring in facilitating the beginning teachers' teaching practice and the improvement that this can bring in the students' learning outcomes, in a developing country's context. The study is largely based on interview data provided by a mentor and two mentee teachers who were the participants of this study. During the study, the mentee teachers disclosed the factors, that they perceived to be the most influential in affecting their progress as teachers. In line with many research studies, this particular research confirms that the attitude and skills of a mentor are a critical factor in raising the confidence and self-esteem of beginning teachers. The findings of the study indicate that mentoring creates a model for extending the learning process and, thus, it evolves to serve a pivotal function for nurturing the beginning teachers' professional development, in their initial years of experience in the classroom. The study pays particular attention to the impact of the mentoring process on improving the students' learning, and suggests a number of strategies for further improving the students' learning outcomes. Based on the results of the research findings, the study confirms that improvement in beginning teachers' teaching practice produces substantial improvement in students' learning outcomes. In addition, some recommendations and implications have been drawn from the study, so as to contribute to the betterment of the mentoring process in Pakistani schools and also in schools in Kyrgyzstan.