A Contrast-Source Inversion-Assisted Attention-Unet for Microwave Imaging

This paper introduces a physics-based intermediate estimate as input to the Attention-Unet (ATTN-Unet) architecture for solving the electromagnetic inverse scattering problem in microwave imaging. This input is calculated from the tenth iteration results of the conventional contrast source inversion...

وصف كامل

محفوظ في:
التفاصيل البيبلوغرافية
المؤلف الرئيسي: Maricar, Mohammed Farook (author)
مؤلفون آخرون: Zakaria, Amer (author), Qaddoumi, Nasser (author)
التنسيق: article
منشور في: 2025
الموضوعات:
الوصول للمادة أونلاين:https://hdl.handle.net/11073/33458
الوسوم: إضافة وسم
لا توجد وسوم, كن أول من يضع وسما على هذه التسجيلة!
_version_ 1870676404178255872
author Maricar, Mohammed Farook
author2 Zakaria, Amer
Qaddoumi, Nasser
author2_role author
author
author_facet Maricar, Mohammed Farook
Zakaria, Amer
Qaddoumi, Nasser
author_role author
dc.creator.none.fl_str_mv Maricar, Mohammed Farook
Zakaria, Amer
Qaddoumi, Nasser
dc.date.none.fl_str_mv 2025-08
2026-06-08T11:06:18Z
2026-06-08T11:06:18Z
dc.format.none.fl_str_mv application/pdf
dc.identifier.none.fl_str_mv M. Farook Maricar, A. Zakaria and N. Qaddoumi, "A Contrast-Source Inversion-Assisted Attention-Unet for Microwave Imaging," in IEEE Access, vol. 13, pp. 142443-142456, 2025, doi: 10.1109/ACCESS.2025.3598475.
2169-3536
https://hdl.handle.net/11073/33458
10.1109/ACCESS.2025.3598475
dc.language.none.fl_str_mv en
dc.publisher.none.fl_str_mv IEEE Xplore
dc.relation.none.fl_str_mv https://doi.org/10.1109/ACCESS.2025.3598475
dc.subject.none.fl_str_mv Deep learning
Convolutional neural network
Inverse scattering
Microwave imaging
Attention Unet
Contrast source inversion
dc.title.none.fl_str_mv A Contrast-Source Inversion-Assisted Attention-Unet for Microwave Imaging
dc.type.none.fl_str_mv Peer-Reviewed
Published version
info:eu-repo/semantics/publishedVersion
info:eu-repo/semantics/article
description This paper introduces a physics-based intermediate estimate as input to the Attention-Unet (ATTN-Unet) architecture for solving the electromagnetic inverse scattering problem in microwave imaging. This input is calculated from the tenth iteration results of the conventional contrast source inversion (CSI) algorithm, which is referred to as ITER10. This input incorporates more physical domain knowledge than the widely used backpropagation (BP) estimate, which corresponds to the zeroth iteration of the CSI. While the non-iterative BP estimate is popular due to its simplicity, this work demonstrates that using ITER10 enhances reconstruction accuracy without significantly increasing computational cost. For comparison and the validation of choosing ITER10, the performance of the ATTN-Unet is evaluated using estimates from other intermediate CSI iterations, namely ITER5, 20, 30, and 40. Further, the network outputs are the reconstructed relative complex permittivity values (real and imaginary) of an imaged object. The networks are tested using synthetic and experimental datasets. The results show that the ITER10-ATTN-Unet significantly enhances reconstruction accuracy, outperforming both the BP-ATTN-Unet and the conventional CSI method. Furthermore, the results demonstrate that the ITER10-ATTN-Unet achieves a better balance between accuracy and computational cost compared to the other ITER-based models. These findings highlight the effectiveness of ITER10 as a strong alternative to BP in improving neural network reconstructions in microwave imaging.
format article
id aus_14d51905ce3563eb64bee432562a2bbd
identifier_str_mv M. Farook Maricar, A. Zakaria and N. Qaddoumi, "A Contrast-Source Inversion-Assisted Attention-Unet for Microwave Imaging," in IEEE Access, vol. 13, pp. 142443-142456, 2025, doi: 10.1109/ACCESS.2025.3598475.
2169-3536
10.1109/ACCESS.2025.3598475
language_invalid_str_mv en
network_acronym_str aus
network_name_str aus
oai_identifier_str oai:repository.aus.edu:11073/33458
publishDate 2025
publisher.none.fl_str_mv IEEE Xplore
repository.mail.fl_str_mv
repository.name.fl_str_mv
repository_id_str
spelling A Contrast-Source Inversion-Assisted Attention-Unet for Microwave ImagingMaricar, Mohammed FarookZakaria, AmerQaddoumi, NasserDeep learningConvolutional neural networkInverse scatteringMicrowave imagingAttention UnetContrast source inversionThis paper introduces a physics-based intermediate estimate as input to the Attention-Unet (ATTN-Unet) architecture for solving the electromagnetic inverse scattering problem in microwave imaging. This input is calculated from the tenth iteration results of the conventional contrast source inversion (CSI) algorithm, which is referred to as ITER10. This input incorporates more physical domain knowledge than the widely used backpropagation (BP) estimate, which corresponds to the zeroth iteration of the CSI. While the non-iterative BP estimate is popular due to its simplicity, this work demonstrates that using ITER10 enhances reconstruction accuracy without significantly increasing computational cost. For comparison and the validation of choosing ITER10, the performance of the ATTN-Unet is evaluated using estimates from other intermediate CSI iterations, namely ITER5, 20, 30, and 40. Further, the network outputs are the reconstructed relative complex permittivity values (real and imaginary) of an imaged object. The networks are tested using synthetic and experimental datasets. The results show that the ITER10-ATTN-Unet significantly enhances reconstruction accuracy, outperforming both the BP-ATTN-Unet and the conventional CSI method. Furthermore, the results demonstrate that the ITER10-ATTN-Unet achieves a better balance between accuracy and computational cost compared to the other ITER-based models. These findings highlight the effectiveness of ITER10 as a strong alternative to BP in improving neural network reconstructions in microwave imaging.American University of SharjahIEEE Xplore2026-06-08T11:06:18Z2026-06-08T11:06:18Z2025-08Peer-ReviewedPublished versioninfo:eu-repo/semantics/publishedVersioninfo:eu-repo/semantics/articleapplication/pdfM. Farook Maricar, A. Zakaria and N. Qaddoumi, "A Contrast-Source Inversion-Assisted Attention-Unet for Microwave Imaging," in IEEE Access, vol. 13, pp. 142443-142456, 2025, doi: 10.1109/ACCESS.2025.3598475.2169-3536https://hdl.handle.net/11073/3345810.1109/ACCESS.2025.3598475enhttps://doi.org/10.1109/ACCESS.2025.3598475oai:repository.aus.edu:11073/334582026-06-09T05:27:58Z
spellingShingle A Contrast-Source Inversion-Assisted Attention-Unet for Microwave Imaging
Maricar, Mohammed Farook
Deep learning
Convolutional neural network
Inverse scattering
Microwave imaging
Attention Unet
Contrast source inversion
status_str publishedVersion
title A Contrast-Source Inversion-Assisted Attention-Unet for Microwave Imaging
title_full A Contrast-Source Inversion-Assisted Attention-Unet for Microwave Imaging
title_fullStr A Contrast-Source Inversion-Assisted Attention-Unet for Microwave Imaging
title_full_unstemmed A Contrast-Source Inversion-Assisted Attention-Unet for Microwave Imaging
title_short A Contrast-Source Inversion-Assisted Attention-Unet for Microwave Imaging
title_sort A Contrast-Source Inversion-Assisted Attention-Unet for Microwave Imaging
topic Deep learning
Convolutional neural network
Inverse scattering
Microwave imaging
Attention Unet
Contrast source inversion
url https://hdl.handle.net/11073/33458