Machine Learning-based X-Ray Projection Interpolation for Improved 4D-CBCT Reconstruction
Respiration-correlated cone-beam computed tomography (4D-CBCT) is an X-ray-based imaging modality that uses reconstruction algorithms to produce time-varying volumetric images of moving anatomy over a cycle of respiratory motion. The quality of the produced images is affected by the number of CBCT p...
محفوظ في:
| المؤلف الرئيسي: | |
|---|---|
| مؤلفون آخرون: | , , |
| التنسيق: | article |
| منشور في: |
2024
|
| الموضوعات: | |
| الوصول للمادة أونلاين: | https://hdl.handle.net/11073/25591 |
| الوسوم: |
إضافة وسم
لا توجد وسوم, كن أول من يضع وسما على هذه التسجيلة!
|
| _version_ | 1864513443357261824 |
|---|---|
| author | Ramesh, Jayroop |
| author2 | Sankalpa, Donthi Mitra, Rohan Dhou, Salam |
| author2_role | author author author |
| author_facet | Ramesh, Jayroop Sankalpa, Donthi Mitra, Rohan Dhou, Salam |
| author_role | author |
| dc.creator.none.fl_str_mv | Ramesh, Jayroop Sankalpa, Donthi Mitra, Rohan Dhou, Salam |
| dc.date.none.fl_str_mv | 2024-09-17T08:13:24Z 2024-09-17T08:13:24Z 2024 |
| dc.format.none.fl_str_mv | application/pdf application/pdf |
| dc.identifier.none.fl_str_mv | J. Ramesh, D. Sankalpa, R. Mitra and S. Dhou, "Machine Learning-based X-Ray Projection Interpolation for Improved 4D-CBCT Reconstruction," in IEEE Open Journal of Engineering in Medicine and Biology, doi: 10.1109/OJEMB.2024.3459622. 2644-1276 https://hdl.handle.net/11073/25591 10.1109/OJEMB.2024.3459622 10.1109/OJEMB.2024.3459622/mm1 |
| dc.language.none.fl_str_mv | en_US |
| dc.publisher.none.fl_str_mv | IEEE |
| dc.relation.none.fl_str_mv | https://ieeexplore.ieee.org/abstract/document/10678916 |
| dc.subject.none.fl_str_mv | 4D-CBCT reconstruction Deep Learning Transfer learning Intermediate Projection Interpolation Multi-output Regression |
| dc.title.none.fl_str_mv | Machine Learning-based X-Ray Projection Interpolation for Improved 4D-CBCT Reconstruction |
| dc.type.none.fl_str_mv | Peer-Reviewed Published version Other info:eu-repo/semantics/publishedVersion info:eu-repo/semantics/article |
| description | Respiration-correlated cone-beam computed tomography (4D-CBCT) is an X-ray-based imaging modality that uses reconstruction algorithms to produce time-varying volumetric images of moving anatomy over a cycle of respiratory motion. The quality of the produced images is affected by the number of CBCT projections available for reconstruction. Interpolation techniques have been used to generate intermediary projections to be used, along with the original projections, for reconstruction. Transfer learning is a powerful approach that harnesses the ability to reuse pre-trained models in solving new problems. Methods: Several state-of-the-art pre-trained deep learning models, used for video frame interpolation, are utilized in this work to generate intermediary projections. Moreover, a novel regression predictive modeling approach is also proposed to achieve the same objective. Digital phantom and clinical datasets are used to evaluate the performance of the models. Results: The results show that the Real-Time Intermediate Flow Estimation (RIFE) algorithm outperforms the others in terms of the Structural Similarity Index Method (SSIM): 0.986 ± 0.010, Peak Signal to Noise Ratio (PSNR): 44.13 ± 2.76, and Mean Square Error (MSE): 18.86 ± 206.90 across all datasets. Moreover, the interpolated projections were used along with the original ones to reconstruct a 4D-CBCT image that was compared to that reconstructed from the original projections only. Conclusions: The reconstructed image using the proposed approach was found to minimize the streaking artifacts, thereby enhancing the image quality. This work demonstrates the advantage of using general-purpose transfer learning algorithms in 4D-CBCT image enhancement. |
| format | article |
| id | aus_e9709203f9cb7e36bbf8a232668ec527 |
| identifier_str_mv | J. Ramesh, D. Sankalpa, R. Mitra and S. Dhou, "Machine Learning-based X-Ray Projection Interpolation for Improved 4D-CBCT Reconstruction," in IEEE Open Journal of Engineering in Medicine and Biology, doi: 10.1109/OJEMB.2024.3459622. 2644-1276 10.1109/OJEMB.2024.3459622 10.1109/OJEMB.2024.3459622/mm1 |
| language_invalid_str_mv | en_US |
| network_acronym_str | aus |
| network_name_str | aus |
| oai_identifier_str | oai:repository.aus.edu:11073/25591 |
| publishDate | 2024 |
| publisher.none.fl_str_mv | IEEE |
| repository.mail.fl_str_mv | |
| repository.name.fl_str_mv | |
| repository_id_str | |
| spelling | Machine Learning-based X-Ray Projection Interpolation for Improved 4D-CBCT ReconstructionRamesh, JayroopSankalpa, DonthiMitra, RohanDhou, Salam4D-CBCT reconstructionDeep LearningTransfer learningIntermediate Projection InterpolationMulti-output RegressionRespiration-correlated cone-beam computed tomography (4D-CBCT) is an X-ray-based imaging modality that uses reconstruction algorithms to produce time-varying volumetric images of moving anatomy over a cycle of respiratory motion. The quality of the produced images is affected by the number of CBCT projections available for reconstruction. Interpolation techniques have been used to generate intermediary projections to be used, along with the original projections, for reconstruction. Transfer learning is a powerful approach that harnesses the ability to reuse pre-trained models in solving new problems. Methods: Several state-of-the-art pre-trained deep learning models, used for video frame interpolation, are utilized in this work to generate intermediary projections. Moreover, a novel regression predictive modeling approach is also proposed to achieve the same objective. Digital phantom and clinical datasets are used to evaluate the performance of the models. Results: The results show that the Real-Time Intermediate Flow Estimation (RIFE) algorithm outperforms the others in terms of the Structural Similarity Index Method (SSIM): 0.986 ± 0.010, Peak Signal to Noise Ratio (PSNR): 44.13 ± 2.76, and Mean Square Error (MSE): 18.86 ± 206.90 across all datasets. Moreover, the interpolated projections were used along with the original ones to reconstruct a 4D-CBCT image that was compared to that reconstructed from the original projections only. Conclusions: The reconstructed image using the proposed approach was found to minimize the streaking artifacts, thereby enhancing the image quality. This work demonstrates the advantage of using general-purpose transfer learning algorithms in 4D-CBCT image enhancement.American University of SharjahIEEE2024-09-17T08:13:24Z2024-09-17T08:13:24Z2024Peer-ReviewedPublished versionOtherinfo:eu-repo/semantics/publishedVersioninfo:eu-repo/semantics/articleapplication/pdfapplication/pdfJ. Ramesh, D. Sankalpa, R. Mitra and S. Dhou, "Machine Learning-based X-Ray Projection Interpolation for Improved 4D-CBCT Reconstruction," in IEEE Open Journal of Engineering in Medicine and Biology, doi: 10.1109/OJEMB.2024.3459622.2644-1276https://hdl.handle.net/11073/2559110.1109/OJEMB.2024.345962210.1109/OJEMB.2024.3459622/mm1en_UShttps://ieeexplore.ieee.org/abstract/document/10678916oai:repository.aus.edu:11073/255912024-09-17T15:58:06Z |
| spellingShingle | Machine Learning-based X-Ray Projection Interpolation for Improved 4D-CBCT Reconstruction Ramesh, Jayroop 4D-CBCT reconstruction Deep Learning Transfer learning Intermediate Projection Interpolation Multi-output Regression |
| status_str | publishedVersion |
| title | Machine Learning-based X-Ray Projection Interpolation for Improved 4D-CBCT Reconstruction |
| title_full | Machine Learning-based X-Ray Projection Interpolation for Improved 4D-CBCT Reconstruction |
| title_fullStr | Machine Learning-based X-Ray Projection Interpolation for Improved 4D-CBCT Reconstruction |
| title_full_unstemmed | Machine Learning-based X-Ray Projection Interpolation for Improved 4D-CBCT Reconstruction |
| title_short | Machine Learning-based X-Ray Projection Interpolation for Improved 4D-CBCT Reconstruction |
| title_sort | Machine Learning-based X-Ray Projection Interpolation for Improved 4D-CBCT Reconstruction |
| topic | 4D-CBCT reconstruction Deep Learning Transfer learning Intermediate Projection Interpolation Multi-output Regression |
| url | https://hdl.handle.net/11073/25591 |