Improving Brain MRI Image Segmentation Quality: A Hybrid Technique for Intensity Inhomogeneity Correction
Intensity inhomogeneity is a significant issue in magnetic resonance imaging (MRI), where the presence of bias field causes distortions in pixel values, resulting in inconsistent and erroneous intensities across the image. This artifact not only hampers accurate diagnosis by radiologists but also ne...
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
FRUCT
2023-11-01
|
| Series: | Proceedings of the XXth Conference of Open Innovations Association FRUCT |
| Subjects: | |
| Online Access: | https://www.fruct.org/publications/volume-34/fruct34/files/Azi.pdf |
| _version_ | 1827381521887002624 |
|---|---|
| author | Samah Ahmed Abdel Aziz Ammar Hawbani Xing-Fu Wang Abdelrahman Samy Talaat Abdelhamid Ismail Maolood Saeed Hamood Alsamhi |
| author_facet | Samah Ahmed Abdel Aziz Ammar Hawbani Xing-Fu Wang Abdelrahman Samy Talaat Abdelhamid Ismail Maolood Saeed Hamood Alsamhi |
| author_sort | Samah Ahmed Abdel Aziz |
| collection | DOAJ |
| description | Intensity inhomogeneity is a significant issue in magnetic resonance imaging (MRI), where the presence of bias field causes distortions in pixel values, resulting in inconsistent and erroneous intensities across the image. This artifact not only hampers accurate diagnosis by radiologists but also negatively impacts the performance of computer-aided diagnosis algorithms, particularly in tasks like segmentation. In our proposed approach, we use a hybrid technique called KIFCM, which integrates K-means and Fuzzy C-means to enhance brain tumor segmentation. K-means provides computational efficiency, while Fuzzy C-means improves accuracy by detecting missed tumor cells. We employ a bias correction method based on the level set framework, removing noise with a median filter and applying the hybrid KIFCM technique for optimal segmentation. Our method effectively addresses intensity variation challenges, ensuring precise brain tumor region segmentation. We compare our results with DFCM and MFFLs, and the comparison shows the efficiency of our proposed method by highlighting the superior quality and accuracy of 81% achieved with requiring less computational time. |
| first_indexed | 2024-03-08T13:56:27Z |
| format | Article |
| id | doaj.art-c5711e1caae642c2bebdace5e8f08166 |
| institution | Directory Open Access Journal |
| issn | 2305-7254 2343-0737 |
| language | English |
| last_indexed | 2024-03-08T13:56:27Z |
| publishDate | 2023-11-01 |
| publisher | FRUCT |
| record_format | Article |
| series | Proceedings of the XXth Conference of Open Innovations Association FRUCT |
| spelling | doaj.art-c5711e1caae642c2bebdace5e8f081662024-01-15T12:32:23ZengFRUCTProceedings of the XXth Conference of Open Innovations Association FRUCT2305-72542343-07372023-11-0134126https://youtu.be/uBV2NotfHbY10.23919/FRUCT60429.2023.10328152Improving Brain MRI Image Segmentation Quality: A Hybrid Technique for Intensity Inhomogeneity CorrectionSamah Ahmed Abdel Aziz0Ammar Hawbani1Xing-Fu Wang2Abdelrahman Samy3Talaat Abdelhamid4Ismail Maolood5Saeed Hamood Alsamhi6University of science ant technology of ChinaUniversity of Science and Technology of ChinaUSTCZagazig UniversityMenoufia UniversityMinistry of Higher Education and Scientific ResearchUniversity of GalwayIntensity inhomogeneity is a significant issue in magnetic resonance imaging (MRI), where the presence of bias field causes distortions in pixel values, resulting in inconsistent and erroneous intensities across the image. This artifact not only hampers accurate diagnosis by radiologists but also negatively impacts the performance of computer-aided diagnosis algorithms, particularly in tasks like segmentation. In our proposed approach, we use a hybrid technique called KIFCM, which integrates K-means and Fuzzy C-means to enhance brain tumor segmentation. K-means provides computational efficiency, while Fuzzy C-means improves accuracy by detecting missed tumor cells. We employ a bias correction method based on the level set framework, removing noise with a median filter and applying the hybrid KIFCM technique for optimal segmentation. Our method effectively addresses intensity variation challenges, ensuring precise brain tumor region segmentation. We compare our results with DFCM and MFFLs, and the comparison shows the efficiency of our proposed method by highlighting the superior quality and accuracy of 81% achieved with requiring less computational time.https://www.fruct.org/publications/volume-34/fruct34/files/Azi.pdfbrain tumormribais correctionintensity inhomogeneity |
| spellingShingle | Samah Ahmed Abdel Aziz Ammar Hawbani Xing-Fu Wang Abdelrahman Samy Talaat Abdelhamid Ismail Maolood Saeed Hamood Alsamhi Improving Brain MRI Image Segmentation Quality: A Hybrid Technique for Intensity Inhomogeneity Correction Proceedings of the XXth Conference of Open Innovations Association FRUCT brain tumor mri bais correction intensity inhomogeneity |
| title | Improving Brain MRI Image Segmentation Quality: A Hybrid Technique for Intensity Inhomogeneity Correction |
| title_full | Improving Brain MRI Image Segmentation Quality: A Hybrid Technique for Intensity Inhomogeneity Correction |
| title_fullStr | Improving Brain MRI Image Segmentation Quality: A Hybrid Technique for Intensity Inhomogeneity Correction |
| title_full_unstemmed | Improving Brain MRI Image Segmentation Quality: A Hybrid Technique for Intensity Inhomogeneity Correction |
| title_short | Improving Brain MRI Image Segmentation Quality: A Hybrid Technique for Intensity Inhomogeneity Correction |
| title_sort | improving brain mri image segmentation quality a hybrid technique for intensity inhomogeneity correction |
| topic | brain tumor mri bais correction intensity inhomogeneity |
| url | https://www.fruct.org/publications/volume-34/fruct34/files/Azi.pdf |
| work_keys_str_mv | AT samahahmedabdelaziz improvingbrainmriimagesegmentationqualityahybridtechniqueforintensityinhomogeneitycorrection AT ammarhawbani improvingbrainmriimagesegmentationqualityahybridtechniqueforintensityinhomogeneitycorrection AT xingfuwang improvingbrainmriimagesegmentationqualityahybridtechniqueforintensityinhomogeneitycorrection AT abdelrahmansamy improvingbrainmriimagesegmentationqualityahybridtechniqueforintensityinhomogeneitycorrection AT talaatabdelhamid improvingbrainmriimagesegmentationqualityahybridtechniqueforintensityinhomogeneitycorrection AT ismailmaolood improvingbrainmriimagesegmentationqualityahybridtechniqueforintensityinhomogeneitycorrection AT saeedhamoodalsamhi improvingbrainmriimagesegmentationqualityahybridtechniqueforintensityinhomogeneitycorrection |