Development of a digital energy modulation framework in projectional radiography

TL;DR

This paper introduces a new framework for x-ray imaging that uses machine learning to translate between different energy domains, showing promising results for improving image quality.

Contribution

The paper presents a novel digital energy modulation framework integrated with machine learning for energy translation in projectional x-ray imaging.

Findings

Machine learning models achieved PSNR of 29.1 ± 2.0 and SSIM of 0.947 ± 0.017 in energy translation tasks.

Translation between polyenergetic domains showed PSNR of 27.4 ± 0.5 and SSIM of 0.909 ± 0.003 in specific projection views.

The results confirm the feasibility of using machine learning for digital energy modulation in x-ray imaging.

Abstract

Digital energy modulation is a novel framework with the potential to enhance projectional x‐ray imaging by enabling translation between different x‐ray energy domains. We evaluate the feasibility of integrating machine learning methods into this approach by leveraging digitally reconstructed radiographs (DRRs) generated from dual‐energy CT datasets.

DRRs were created in 15° increments from 0° to 90°, producing 3500 images per energy domain (2 polyenergetic, 4 monoenergetic). A supervised deep‐learning approach was used to train models for energy translation, focusing on conversions between polyenergetic domains and from polyenergetic to monoenergetic images. Model performance was assessed using peak signal‐to‐noise ratio (PSNR), structural similarity index (SSIM), mean squared error (MSE), and mean absolute percentage error (MAPE). Cross‐validation and projection‐specific dataset splits were used for evaluation.

The models trained using cross‐validation on the various energy translations achieved the following results: PSNR: 29.1 ± 2.0, SSIM: 0.947 ± 0.017, MSE: 169.1 ± 68.3, MAPE: 8.2% ± 1.8%. When translating between polyenergetic high‐energy and low‐energy domains in projection‐specific datasets (anterior‐posterior [0°] and lateral [90°] views), models achieved the following results: PSNR: 27.4 ± 0.5, SSIM: 0.909 ± 0.003, MSE: 195.9 ± 39.7, MAPE: 10.4% ± 2.1%.

These findings demonstrate the feasibility of a digital energy modulation framework for projectional x‐ray imaging using machine learning for energy translation. The results support the potential of this approach to enhance projectional x‐ray imaging, though future work is needed to refine the models and further explore clinical applications.

Figures