(For USM Staff/Student Only)

EngLib USM > Ω School of Electrical & Electronic Engineering >

Graph cut technique of mri brain segmentation/Muhammad Rezza Abdul Rahim

Graph cut technique of mri brain segmentation_Muhammad Rezza Abdul Rahim_E3_2011_NI
Segmentasi tepat dan cekap boleh diaplikasikan dalam bidang sains perubatan dan klinikal bagi membezakan jenis-jenis tisu dalam imej MRI untuk otak. Kebanyakan gambar otak mengandungi hingar dan ketidakhomogenasi menyebabkan tugas segmentasi tepat menjadi sukar. Graph Cuts adalah algoritma baru dan semakin popular yang diintegrasikan dalam penyelesaian masalah penglihatan terutamanya pada peringkat rendah seperti segmentasi imej, pembinaan objek, pemuliharaan imej dan anggaran perbezaan. Dalam kajian ini, teknik perkomputeran Graph Cuts dieksplorasi untuk menilai kecekapannya dalam segmentasi tisu dalam imej MRI dibandingkan dengan segmentasi Graph Cuts secara manual. Kaedah yang dicadangkan adalah segmentasi automatik dengan menggunakan susunan data yang berbeza bagi imej otak. Untuk memudahkan proses, pemalar yang sesuai ditetapkan dalam parameter Graph Cuts setelah kajian dilaksanakan kepada imej yang tertentu. Kit yang dicadangkan untuk segmentasi MRI otak secara automatik mempunyai kemampuan untuk menjalankan proses segmentasi dengan lebih cekap serta membezakan jenis-jenis tisu mengikut perbezaan keamatan yang lebih tepat berbanding teknik Graph Cuts secara manual. Hasil kajian membuktikan bahawa Graph Cuts amat sesuai untuk mengklasifikasikan pixel dan memberikan analisis kuantitatif untuk otak. ___________________________________________________________________________________ Accurate and efficient brain image segmentation is the basis for many applications in clinical for scientific and clinical researchers to distinguish tissue types in brain MRI. As brain image mostly contain noise, inhomogeneity and sometimes deviation, hence, accurate segmentation of the brain image is a very difficult task. Graph cut is a new algorithm currently being integrated in the computer vision community. The graph cut is popular in solving many low level vision problems such as image segmentation, object reconstruction, and image restoration. In this study, it has been explored to evaluate its computerized efficiency to segment the MRI brain tissue compare with the manual graph cut segmentation. Further extensions of the proposed automated graph cut segmentation are being done to achieve the quantity evaluation of tissue classification using different data of brain images. For robustness, the most suitable constants are fixed as the graph cut parameters after experiments have been done to the images. The toolkit proposed for graph cut technique of MRI segmentation has the capability in executing the segmentation faster and classify the different types of tissue according to the intensity difference much sharper than the manual graph cut method. These segmentation results proved that graph cut technique is suitable to be classifying the tissues type and provide quantitative analysis of brain volume.
Contributor(s):
Muhammad Rezza Abdul Rahim - Author
Language:
English
Subject Keywords:
brain ; noise ; graph
First presented to the public:
4/1/2011
Original Publication Date:
3/4/2020
Previously Published By:
Universiti Sains Malaysia
Place Of Publication:
School of Electrical & Electronic Engineering
Citation:
Extents:
Number of Pages - 107
License Grantor / Date Granted:
  / ( View License )
Date Deposited
2020-03-04 14:26:54.443
Submitter:
Nor Hayati Ismail

All Versions

Thumbnail Name Version Created Date
Graph cut technique of mri brain segmentation/Muhammad Rezza Abdul Rahim1 2020-03-04 14:26:54.443