Assessment of Virtual Grid software for improving X-ray image quality and reducing radiation dose

Abstract

Background and rationale: X-ray imaging is routinely used for diagnostic purposes. Scattered radiation from the patient is a major image degradation factor, as scattered X-rays appear as mislocated events and are undesirable since it increases the patient dose. The conventional method of reducing scattered radiation is an anti-scatter physical grid (PG) which is placed directly between the patient and the detector to remove or absorb scattered radiation whilst allowing primary x-ray radiation to pass to the image receptor, but this increases the dose and requires precise positioning and alignment. Recently, a new ‘Virtual Grid’ image processing software (VG) has been developed to correct for scattered X-rays and overcome some of the technical issues associated with grid devices. Objectives: This PhD project aims to assess the use and effectiveness of VG software on radiographs acquired without a physical grid in terms of image quality and the image quality and radiation dose in comparison to physical grid radiographs. Materials and methods: Initially a scoping review was conducted to identify and highlight published papers regarding X-ray scatter correction software for X-ray imaging. The Leeds model TOR CDR® phantom with polymethylmethacrylate image quality and an anatomical anthropomorphic PBU-50 whole-body phantom with added fabricated substitute layers of fat and lean tissue were experimentally assessed Clinically, retrospective clinical patient data was examined. These data were used to compare image quality and radiation dose of Gridless, VG software and PG approaches. Results: VG software resulted in an increase in image quality objectively and subjectively compared to raw gridless radiographs. However, PG images generally had higher image quality than VG software generated images, especially in clinical practice. The radiation dose in terms of DAP and the effective dose of Gridless with VG software procedures was lower than PG. VG software shows potential under some clinical conditions such as patients who are critically ill in bed or cannot stand, and when PG can be challenging or not appropriate for large body regions. Conclusion: This PhD project has demonstrated the potential of the VG software in terms of improving the image quality of the Gridless radiography. VG software promises to lower radiation dose levels in terms of DAP and effective dose compared to the conventional PG approach. This research contributes to research on the potential of scatter correction software in improving image quality and reducing radiation dose. However, it has limitations and further research is needed.