Evaluation of a Generative Adversarial Network to Improve Image Quality and Reduce Radiation-Dose during Digital Breast Tomosynthesis
In this study, we evaluated the improvement of image quality in digital breast tomosynthesis under low-radiation dose conditions of pre-reconstruction processing using conditional generative adversarial networks [cGAN (pix2pix)]. Pix2pix pre-reconstruction processing with filtered back projection (F...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-02-01
|
| Series: | Diagnostics |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2075-4418/12/2/495 |
| _version_ | 1827655637295693824 |
|---|---|
| author | Tsutomu Gomi Yukie Kijima Takayuki Kobayashi Yukio Koibuchi |
| author_facet | Tsutomu Gomi Yukie Kijima Takayuki Kobayashi Yukio Koibuchi |
| author_sort | Tsutomu Gomi |
| collection | DOAJ |
| description | In this study, we evaluated the improvement of image quality in digital breast tomosynthesis under low-radiation dose conditions of pre-reconstruction processing using conditional generative adversarial networks [cGAN (pix2pix)]. Pix2pix pre-reconstruction processing with filtered back projection (FBP) was compared with and without multiscale bilateral filtering (MSBF) during pre-reconstruction processing. Noise reduction and preserve contrast rates were compared using full width at half-maximum (FWHM), contrast-to-noise ratio (CNR), peak signal-to-noise ratio (PSNR), and structural similarity (SSIM) in the in-focus plane using a BR3D phantom at various radiation doses [reference-dose (automatic exposure control reference dose: AECrd), 50% and 75% reduction of AECrd] and phantom thicknesses (40 mm, 50 mm, and 60 mm). The overall performance of pix2pix pre-reconstruction processing was effective in terms of FWHM, PSNR, and SSIM. At ~50% radiation-dose reduction, FWHM yielded good results independently of the microcalcification size used in the BR3D phantom, and good noise reduction and preserved contrast. PSNR results showed that pix2pix pre-reconstruction processing represented the minimum in the error with reference FBP images at an approximately 50% reduction in radiation-dose. SSIM analysis indicated that pix2pix pre-reconstruction processing yielded superior similarity when compared with and without MSBF pre-reconstruction processing at ~50% radiation-dose reduction, with features most similar to the reference FBP images. Thus, pix2pix pre-reconstruction processing is promising for reducing noise with preserve contrast and radiation-dose reduction in clinical practice. |
| first_indexed | 2024-03-09T22:10:26Z |
| format | Article |
| id | doaj.art-7c40da3217914925bec7f6092706a74a |
| institution | Directory Open Access Journal |
| issn | 2075-4418 |
| language | English |
| last_indexed | 2024-03-09T22:10:26Z |
| publishDate | 2022-02-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Diagnostics |
| spelling | doaj.art-7c40da3217914925bec7f6092706a74a2023-11-23T19:33:01ZengMDPI AGDiagnostics2075-44182022-02-0112249510.3390/diagnostics12020495Evaluation of a Generative Adversarial Network to Improve Image Quality and Reduce Radiation-Dose during Digital Breast TomosynthesisTsutomu Gomi0Yukie Kijima1Takayuki Kobayashi2Yukio Koibuchi3School of Allied Health Sciences, Kitasato University, Sagamihara 252-0373, Kanagawa, JapanDepartment of Radiology, National Hospital Organization Takasaki General Medical Center, Takasaki 370-0829, Gunma, JapanDepartment of Radiology, Kitasato University Kitasato Institute Hospital, Shirokane, Minato-ku, Tokyo 108-8642, JapanDepartment of Breast and Endocrine Surgery, National Hospital Organization Takasaki General Medical Center, Takasaki 370-0829, Gunma, JapanIn this study, we evaluated the improvement of image quality in digital breast tomosynthesis under low-radiation dose conditions of pre-reconstruction processing using conditional generative adversarial networks [cGAN (pix2pix)]. Pix2pix pre-reconstruction processing with filtered back projection (FBP) was compared with and without multiscale bilateral filtering (MSBF) during pre-reconstruction processing. Noise reduction and preserve contrast rates were compared using full width at half-maximum (FWHM), contrast-to-noise ratio (CNR), peak signal-to-noise ratio (PSNR), and structural similarity (SSIM) in the in-focus plane using a BR3D phantom at various radiation doses [reference-dose (automatic exposure control reference dose: AECrd), 50% and 75% reduction of AECrd] and phantom thicknesses (40 mm, 50 mm, and 60 mm). The overall performance of pix2pix pre-reconstruction processing was effective in terms of FWHM, PSNR, and SSIM. At ~50% radiation-dose reduction, FWHM yielded good results independently of the microcalcification size used in the BR3D phantom, and good noise reduction and preserved contrast. PSNR results showed that pix2pix pre-reconstruction processing represented the minimum in the error with reference FBP images at an approximately 50% reduction in radiation-dose. SSIM analysis indicated that pix2pix pre-reconstruction processing yielded superior similarity when compared with and without MSBF pre-reconstruction processing at ~50% radiation-dose reduction, with features most similar to the reference FBP images. Thus, pix2pix pre-reconstruction processing is promising for reducing noise with preserve contrast and radiation-dose reduction in clinical practice.https://www.mdpi.com/2075-4418/12/2/495digital breast tomosynthesisgenerative adversarial networksradiation-dose reductionimprove image quality |
| spellingShingle | Tsutomu Gomi Yukie Kijima Takayuki Kobayashi Yukio Koibuchi Evaluation of a Generative Adversarial Network to Improve Image Quality and Reduce Radiation-Dose during Digital Breast Tomosynthesis Diagnostics digital breast tomosynthesis generative adversarial networks radiation-dose reduction improve image quality |
| title | Evaluation of a Generative Adversarial Network to Improve Image Quality and Reduce Radiation-Dose during Digital Breast Tomosynthesis |
| title_full | Evaluation of a Generative Adversarial Network to Improve Image Quality and Reduce Radiation-Dose during Digital Breast Tomosynthesis |
| title_fullStr | Evaluation of a Generative Adversarial Network to Improve Image Quality and Reduce Radiation-Dose during Digital Breast Tomosynthesis |
| title_full_unstemmed | Evaluation of a Generative Adversarial Network to Improve Image Quality and Reduce Radiation-Dose during Digital Breast Tomosynthesis |
| title_short | Evaluation of a Generative Adversarial Network to Improve Image Quality and Reduce Radiation-Dose during Digital Breast Tomosynthesis |
| title_sort | evaluation of a generative adversarial network to improve image quality and reduce radiation dose during digital breast tomosynthesis |
| topic | digital breast tomosynthesis generative adversarial networks radiation-dose reduction improve image quality |
| url | https://www.mdpi.com/2075-4418/12/2/495 |
| work_keys_str_mv | AT tsutomugomi evaluationofagenerativeadversarialnetworktoimproveimagequalityandreduceradiationdoseduringdigitalbreasttomosynthesis AT yukiekijima evaluationofagenerativeadversarialnetworktoimproveimagequalityandreduceradiationdoseduringdigitalbreasttomosynthesis AT takayukikobayashi evaluationofagenerativeadversarialnetworktoimproveimagequalityandreduceradiationdoseduringdigitalbreasttomosynthesis AT yukiokoibuchi evaluationofagenerativeadversarialnetworktoimproveimagequalityandreduceradiationdoseduringdigitalbreasttomosynthesis |