A comparison of three Deformable Image Registration Algorithms in 4DCT using conventional contour based methods and voxel-by-voxel comparison methods.
Background: Commonly used methods of assessing the accuracy of Deformable Image Registration (DIR) rely on image segmentation or landmark selection. These methods are very labor intensive and thus limited to relatively small number of image pairs. The direct voxel-by-voxel comparison can be automate...
| Main Authors: | , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Frontiers Media S.A.
2015-02-01
|
| Series: | Frontiers in Oncology |
| Subjects: | |
| Online Access: | http://journal.frontiersin.org/Journal/10.3389/fonc.2015.00017/full |
| _version_ | 1819238671970205696 |
|---|---|
| author | Mirek eFatyga Nesrin eDogan Jeffrey eWilliamson Elizabeth eWeiss William eSleeman William eLehman Baoshe eZhang Krishni eWijesooriya Gary eChristensen |
| author_facet | Mirek eFatyga Nesrin eDogan Jeffrey eWilliamson Elizabeth eWeiss William eSleeman William eLehman Baoshe eZhang Krishni eWijesooriya Gary eChristensen |
| author_sort | Mirek eFatyga |
| collection | DOAJ |
| description | Background: Commonly used methods of assessing the accuracy of Deformable Image Registration (DIR) rely on image segmentation or landmark selection. These methods are very labor intensive and thus limited to relatively small number of image pairs. The direct voxel-by-voxel comparison can be automated to examine fluctuations in DIR quality on a long series of image pairs.Methods: A voxel-by-voxel comparison of three DIR algorithms applied to lung patients is presented. Registrations are compared by comparing volume histograms formed both with individual DIR maps and with a voxel-by-voxel subtraction of the two maps. When two DIR maps agree one concludes that both maps are interchangeable in treatment planning applications, though one cannot conclude that either one agrees with the ground truth. If two DIR maps significantly disagree one concludes that at least one of the maps deviates from the ground truth. We use the method to compare three DIR algorithms applied to peak inhale-peak exhale registrations of 4DFBCT data obtained from thirteen patients. Results: All three algorithms appear to be nearly equivalent when compared using DICE similarity coefficients. A comparison based on Jacobian Volume Histograms shows that all three algorithms measure changes in total volume of the lungs with reasonable accuracy, but show large differences in the variance of Jacobian distribution on all contoured structures. Analysis of voxel-by-voxel subtraction of DIR maps shows that the three algorithms differ to a degree which is sufficient to create a potential for dosimetric discrepancy during dose accumulation.Conclusions: DIR algorithms can perform well in some clinical applications, while potentially fail in others. These algorithms are best treated as potentially useful approximations of tissue deformation that need to be separately validated for every intended clinical application. |
| first_indexed | 2024-12-23T13:39:56Z |
| format | Article |
| id | doaj.art-bb5e3e1c9fc847e092913cd2347648ec |
| institution | Directory Open Access Journal |
| issn | 2234-943X |
| language | English |
| last_indexed | 2024-12-23T13:39:56Z |
| publishDate | 2015-02-01 |
| publisher | Frontiers Media S.A. |
| record_format | Article |
| series | Frontiers in Oncology |
| spelling | doaj.art-bb5e3e1c9fc847e092913cd2347648ec2022-12-21T17:44:54ZengFrontiers Media S.A.Frontiers in Oncology2234-943X2015-02-01510.3389/fonc.2015.00017119846A comparison of three Deformable Image Registration Algorithms in 4DCT using conventional contour based methods and voxel-by-voxel comparison methods.Mirek eFatyga0Nesrin eDogan1Jeffrey eWilliamson2Elizabeth eWeiss3William eSleeman4William eLehman5Baoshe eZhang6Krishni eWijesooriya7Gary eChristensen8Mayo Clinic AZUniversity of MiamiVirginia Commonwealth UniversityVirginia Commonwealth UniversityVirginia Commonwealth UniversityVirginia Commonwealth UniversityVirginia Commonwealth UniversityUniversity of Virginia Health SystemsUniversity of IowaBackground: Commonly used methods of assessing the accuracy of Deformable Image Registration (DIR) rely on image segmentation or landmark selection. These methods are very labor intensive and thus limited to relatively small number of image pairs. The direct voxel-by-voxel comparison can be automated to examine fluctuations in DIR quality on a long series of image pairs.Methods: A voxel-by-voxel comparison of three DIR algorithms applied to lung patients is presented. Registrations are compared by comparing volume histograms formed both with individual DIR maps and with a voxel-by-voxel subtraction of the two maps. When two DIR maps agree one concludes that both maps are interchangeable in treatment planning applications, though one cannot conclude that either one agrees with the ground truth. If two DIR maps significantly disagree one concludes that at least one of the maps deviates from the ground truth. We use the method to compare three DIR algorithms applied to peak inhale-peak exhale registrations of 4DFBCT data obtained from thirteen patients. Results: All three algorithms appear to be nearly equivalent when compared using DICE similarity coefficients. A comparison based on Jacobian Volume Histograms shows that all three algorithms measure changes in total volume of the lungs with reasonable accuracy, but show large differences in the variance of Jacobian distribution on all contoured structures. Analysis of voxel-by-voxel subtraction of DIR maps shows that the three algorithms differ to a degree which is sufficient to create a potential for dosimetric discrepancy during dose accumulation.Conclusions: DIR algorithms can perform well in some clinical applications, while potentially fail in others. These algorithms are best treated as potentially useful approximations of tissue deformation that need to be separately validated for every intended clinical application.http://journal.frontiersin.org/Journal/10.3389/fonc.2015.00017/fullmedical imagingDeformable Image RegistrationDeformable Dose Additionadiation oncology4DCT |
| spellingShingle | Mirek eFatyga Nesrin eDogan Jeffrey eWilliamson Elizabeth eWeiss William eSleeman William eLehman Baoshe eZhang Krishni eWijesooriya Gary eChristensen A comparison of three Deformable Image Registration Algorithms in 4DCT using conventional contour based methods and voxel-by-voxel comparison methods. Frontiers in Oncology medical imaging Deformable Image Registration Deformable Dose Addition adiation oncology 4DCT |
| title | A comparison of three Deformable Image Registration Algorithms in 4DCT using conventional contour based methods and voxel-by-voxel comparison methods. |
| title_full | A comparison of three Deformable Image Registration Algorithms in 4DCT using conventional contour based methods and voxel-by-voxel comparison methods. |
| title_fullStr | A comparison of three Deformable Image Registration Algorithms in 4DCT using conventional contour based methods and voxel-by-voxel comparison methods. |
| title_full_unstemmed | A comparison of three Deformable Image Registration Algorithms in 4DCT using conventional contour based methods and voxel-by-voxel comparison methods. |
| title_short | A comparison of three Deformable Image Registration Algorithms in 4DCT using conventional contour based methods and voxel-by-voxel comparison methods. |
| title_sort | comparison of three deformable image registration algorithms in 4dct using conventional contour based methods and voxel by voxel comparison methods |
| topic | medical imaging Deformable Image Registration Deformable Dose Addition adiation oncology 4DCT |
| url | http://journal.frontiersin.org/Journal/10.3389/fonc.2015.00017/full |
| work_keys_str_mv | AT mirekefatyga acomparisonofthreedeformableimageregistrationalgorithmsin4dctusingconventionalcontourbasedmethodsandvoxelbyvoxelcomparisonmethods AT nesrinedogan acomparisonofthreedeformableimageregistrationalgorithmsin4dctusingconventionalcontourbasedmethodsandvoxelbyvoxelcomparisonmethods AT jeffreyewilliamson acomparisonofthreedeformableimageregistrationalgorithmsin4dctusingconventionalcontourbasedmethodsandvoxelbyvoxelcomparisonmethods AT elizabetheweiss acomparisonofthreedeformableimageregistrationalgorithmsin4dctusingconventionalcontourbasedmethodsandvoxelbyvoxelcomparisonmethods AT williamesleeman acomparisonofthreedeformableimageregistrationalgorithmsin4dctusingconventionalcontourbasedmethodsandvoxelbyvoxelcomparisonmethods AT williamelehman acomparisonofthreedeformableimageregistrationalgorithmsin4dctusingconventionalcontourbasedmethodsandvoxelbyvoxelcomparisonmethods AT baosheezhang acomparisonofthreedeformableimageregistrationalgorithmsin4dctusingconventionalcontourbasedmethodsandvoxelbyvoxelcomparisonmethods AT krishniewijesooriya acomparisonofthreedeformableimageregistrationalgorithmsin4dctusingconventionalcontourbasedmethodsandvoxelbyvoxelcomparisonmethods AT garyechristensen acomparisonofthreedeformableimageregistrationalgorithmsin4dctusingconventionalcontourbasedmethodsandvoxelbyvoxelcomparisonmethods AT mirekefatyga comparisonofthreedeformableimageregistrationalgorithmsin4dctusingconventionalcontourbasedmethodsandvoxelbyvoxelcomparisonmethods AT nesrinedogan comparisonofthreedeformableimageregistrationalgorithmsin4dctusingconventionalcontourbasedmethodsandvoxelbyvoxelcomparisonmethods AT jeffreyewilliamson comparisonofthreedeformableimageregistrationalgorithmsin4dctusingconventionalcontourbasedmethodsandvoxelbyvoxelcomparisonmethods AT elizabetheweiss comparisonofthreedeformableimageregistrationalgorithmsin4dctusingconventionalcontourbasedmethodsandvoxelbyvoxelcomparisonmethods AT williamesleeman comparisonofthreedeformableimageregistrationalgorithmsin4dctusingconventionalcontourbasedmethodsandvoxelbyvoxelcomparisonmethods AT williamelehman comparisonofthreedeformableimageregistrationalgorithmsin4dctusingconventionalcontourbasedmethodsandvoxelbyvoxelcomparisonmethods AT baosheezhang comparisonofthreedeformableimageregistrationalgorithmsin4dctusingconventionalcontourbasedmethodsandvoxelbyvoxelcomparisonmethods AT krishniewijesooriya comparisonofthreedeformableimageregistrationalgorithmsin4dctusingconventionalcontourbasedmethodsandvoxelbyvoxelcomparisonmethods AT garyechristensen comparisonofthreedeformableimageregistrationalgorithmsin4dctusingconventionalcontourbasedmethodsandvoxelbyvoxelcomparisonmethods |