Projection radiography in obese populations: how can image quality be improved?

Abstract

Introduction The prevalence of obesity is increasing globally; this, in turn, has impacted negatively on health service delivery, especially in radiology. The high radiation dose is one of the main challenges as the literature suggested. Objectives This thesis aimed to; identify the most challenging procedures in projection radiography, quantify the FM and FFM in these areas in order to build obese phantoms, to build multiple phantoms, which then were used to investigate dose optimisation in obese patients. Materials and methods Radiation dose delivered to this group of patients in the UK was assessed using DAP. Based on the result of that, the FM and FFM of the most challenging areas were then quantified based on DXA scans of 264 female participants. The phantoms were then built using PMC 121/30 dry with other additives( CaCo3, phenolic microspheres) Density, HU and attenuation coefficient were used to validate the phantoms materials. Dose optimisation was then carried out using factorial design. Results Abdominal and lumbar spine radiographs were the most challenging body parts with over 600% increase in DAP compared to the UK NDRL. The DXA scans show an increase in FFM along with FM and prediction models were produced to predict FFM and FM in these areas. The phantoms were built, and all validations steps show agreement with the ICRU 44 report. kVp and filtration were the most beneficial factors in dose optimisation in this thesis. kVp and filtration were the most influential factors in dose optimisation. mAs prediction model was also developed. Conclusions High doses are delivered to obese patients primarily in the lumbar and abdominal radiograph. Subsequently, high chance of induced cancer is expected. The DXA data shows a big gap in the literature where previously constructed phantoms are considering FM only and ignoring the FFM. The prediction models produced will help standardised construction of obese computational and physical phantoms. The phantoms built show a promising pathway in producing obese phantom with different BMI to tackle the dose and image quality issues. Low kVp with high filtration is recommended when imaging this group of patients. mAs prediction model can be used and the tube limit should be known to act accordingly.