EUVIMEDEUVIMED entdecken
Crossref

A deep-learning-based scatter correction with water equivalent path length map for digital radiography

Masayuki Hattori, Hisato Tsubakiya, Sung-Hyun Lee, Takayuki Kanai, Koji Suzuki, Tetsuya Yuasa

Radiological Physics and Technology · 2024

DOI: 10.1007/s12194-024-00807-9 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Radiological Physics and Technology, 2024. Verfasst von Masayuki Hattori und 5 weiteren Autor:innen. Beteiligte Einrichtung: Graduate School of Science and Engineering, Yamagata University, Yonezawa, 992-8510, Japan. m-hattori@med.id.yamagata-u.ac.jp. Hauptthemen laut MEDLINE: Deep Learning, Scattering, Radiation, Water, Phantoms, Imaging. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 39 Literaturangaben.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

We proposed a new deep learning (DL) model for accurate scatter correction in digital radiography. The proposed network featured a pixel-wise water equivalent path length (WEPL) map of subjects with diverse sizes and 3D inner structures. The proposed U-Net model comprises two concatenated modules: one for generating a WEPL map and the other for predicting scatter using the WEPL map as auxiliary information. First, 3D CT images were used as numerical phantoms for training and validation, generating observed and scattered images by Monte Carlo simulation, and WEPL maps using Siddon's algorithm. Then, we optimised the model without overfitting. Next, we validated the proposed model's performance by comparing it with other DL models. The proposed model obtained scatter-corrected images with a peak signal-to-noise ratio of 44.24 ± 2.89 dB and a structural similarity index measure of 0.9987 ± 0.0004, which were higher than other DL models. Finally, scatter fractions (SFs) were compared with other DL models using an actual phantom to confirm practicality. Among DL models, the proposed model showed the smallest deviation from measured SF values. Furthermore, using an actual radiograph containing an acrylic object, the contrast-to-noise ratio (CNR) of the proposed model and the anti-scatter grid were compared. The CNR of the images corrected using the proposed model are 16% and 82% higher than those of the raw and grid-applied images, respectively. The advantage of the proposed method is that no actual radiography system is required for collecting training dataset, as the dataset is created from CT images using Monte Carlo simulation.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Deep LearningScattering, RadiationWaterPhantoms, ImagingRadiographic Image EnhancementMonte Carlo MethodImage Processing, Computer-AssistedHumansTomography, X-Ray ComputedAlgorithmsSignal-To-Noise RatioImaging, Three-Dimensional

Stichwörter der Autor:innen

Deep learningDigital radiographyMonte Carlo simulationScatter correctionU-NetWater equivalent path length

Substanzen

Water

Autor:innen und Institutionen

  • Graduate School of Science and Engineering, Yamagata University, Yonezawa, 992-8510, Japan. m-hattori@med.id.yamagata-u.ac.jp.
  • Department of Radiology, Yamagata University Hospital, Yamagata, 990-9585, Japan. m-hattori@med.id.yamagata-u.ac.jp.
  • Graduate School of Science and Engineering, Yamagata University, Yonezawa, 992-8510, Japan.
  • Department of Heavy Particle Medical Science, Graduate School of Medicine, Yamagata University, Yamagata, 990-9585, Japan.
  • Department of Radiation Oncology, Tokyo Women's Medical University, Shinjuku, Tokyo, Japan.
  • Department of Radiology, Yamagata University Hospital, Yamagata, 990-9585, Japan.

Zitieren

APA

Hattori, M., Tsubakiya, H., Lee, S. H., Kanai, T., Suzuki, K., & Yuasa, T. (2024). A deep-learning-based scatter correction with water equivalent path length map for digital radiography. Radiological Physics and Technology. https://doi.org/10.1007/s12194-024-00807-9

Vancouver

Hattori M, Tsubakiya H, Lee SH, Kanai T, Suzuki K, Yuasa T. A deep-learning-based scatter correction with water equivalent path length map for digital radiography. Radiological Physics and Technology. 2024. doi:10.1007/s12194-024-00807-9

BibTeX
@article{Hattori2024,
  author = {Hattori, Masayuki and Tsubakiya, Hisato and Lee, Sung-Hyun and Kanai, Takayuki and Suzuki, Koji and Yuasa, Tetsuya},
  title = {A deep-learning-based scatter correction with water equivalent path length map for digital radiography},
  journal = {Radiological Physics and Technology},
  year = {2024},
  doi = {10.1007/s12194-024-00807-9}
}
Kostenlos starten →