Establishment and optimisation of diagnostic reference levels for common CT examinations in Nigeria

Abstract

Purpose: To investigate CT radiation doses in Nigeria and establish DRLs based on CTDIvol, DLP and SSDE and propose a methodology for dose optimisation using a QA phantom and a human cadaver prior to clinical implementation. Materials and methods: The study was carried out in two phases. Phase I involved a survey of CT dose indices including scan parameters and patient demographic information in 23 CT scanners listed in the NNRA registry in Nigeria. Phase II focused on dose optimisation which was divided into three stages. Stage I: Manipulation of the scan parameters and monitoring their effect on the dose indices and psycho-physical parameters using a GE QA phantom. Stage II: Application of the established QA phantom protocols on the human cadaver with further optimisation which was followed with quantitative image analysis based on SNR and CNR, and qualitative evaluation of result of clinical images based on VGA, VGC and VGR. Stage III: clinical implementation of the optimised protocols and re-establishment of DRLs to monitor the effect of optimisation. Results: CT DRLs were established for adults and paediatrics. Fifteen adult QA optimised protocols were established in five CT scanners. Of the 15 QA phantom protocols, nine protocols from three scanners were tested using a cadaver. Of the nine protocols, six were further optimised, two could not be further optimised, whilst, one was not accepted as it did not satisfy image quality requirements. Furthermore, two of the six cadaver optimised protocols were not implemented due to the radiologists’ preference of image quality, opting for the established QA optimised protocols. The cadaver optimised protocols were implemented in two of the three CT scanners. Dose reductions achieved for the Toshiba scanner were 37% and 47% for the brain CTDIvol and DLP respectively, whilst, for chest 82%, 81% and 84%; and for abdomen 50%, 70% and 45% based on the CTDIvol, DLP and SSDE respectively. Meanwhile, for the GE 16-slice, the dose was reduced by 71% and 64% for the brain CTDIvol and DLP respectively, whilst, for chest by 17%, 3% and 16%; and for abdomen by 8%, 16% and 10% based on the CTDIvol, DLP and SSDE respectively. Conclusion: The established DRLs identified CT scanners in Nigeria with higher doses. This guided the implementation of optimised protocols in selected CT scanners providing acceptable image quality based on VGR and quantitative evaluation in terms of SNR and CNR.