CT Dosimetry : What Has Been Achieved and What Remains to Be Done
John Damilakis, PhD. Investigative Radiology Vol 56, N°1, January 2021
Fibermetrix has prepared a summary of the article “CT Dosimetry, What Has Been Achieved and What Remains to Be Done” published in Investigative Radiology.
Why is there a need for increased accuracy in measuring patient dose in CT?
A global need
Computed Tomography (CT) is the largest source of exposure in medical imaging and its share of the collective dose has increased in recent years (63% in 2016 vs. 50% in 2006 in the United States). This trend is also observed in several European countries. It is difficult to predict the evolution of doses delivered in CT in the coming years because, on the one hand, advances in dose reduction tools, image reconstruction software, detector technology and dose management systems may lead to a reduction in patient dose. On the other hand, there will potentially be an increase in the use of new techniques and protocols such as CT perfusions, use of CT screening for colon and lung cancer, use of dual energy scanners or hybrid SPECT and PET CT, which may lead to an increase in procedures and associated dose. Accurate radiation dose estimation is therefore essential for the justification of exposure, optimization and management of the dose delivered to patients in CT. Dose measurements are performed to assess CT scan performance (commissioning and quality control); to compare and optimize CT techniques, procedures or technologies; to establish DRLs; to provide information for CT procedures justification; to assess radiation doses and associated risks by evaluating effective dose per procedure, collective effective dose, annual average individual effective dose and other parameters in radiation surveys and epidemiological studies.
Limited dose indicators
The most frequently used metrics for characterizing CT radiation dose are Computed Tomography Dose Index (CTDI) and the Dose Length Product (DLP), which represent an average absorbed dose along the Z-axis in standardized cylindrical PMMA phantoms 14cm long and 16cm diameter for skull examinations and 32cm for body examinations. However, these indicators have major limitations because they do not represent the average dose received by the patient due to the obvious differences in composition/morphology between PMMA phantoms and a patient and less scattered radiation. In addition, there are uncertainties regarding the phantom diameters to be used for neck dose calculations in pediatrics and for whole body examinations in adults and the 10cm ionization chambers used to determine the CTDI are not sized to accurately measure the dose on fields >40mm often found on the latest generation of scanners.
Why do we need the dose uncertainty to be as small as possible?
First, the health impact of radiation has been focused on stochastic effects and in particular on cancer, although there is compelling evidence of non-cancer effects, such as lens opacity and cataracts, and cardiovascular disease at relatively low doses. Second, a considerable number of patients undergo multiple examinations. These patients are sometimes exposed to an effective dose in excess of the 100mSv threshold often considered for significant stochastic risk. Underestimated or overestimated patient dosimetry leads to incorrect priorities for medical radiation protection research and incorrect dose management of these patients. Third, dosimetry is critical when pregnant patients are scheduled for CT scanning. Specific dose measurements and accurate estimation of the dose to the embryo/fetus may be required above 10mGy to the unborn child and abortion may be considered if this dose is above 100 mGy. Dosimetry must therefore be accurate. Fourth, high accuracy is needed for the dose measurements used to compare different techniques. Fifth, it is important that the uncertainty of the dose measurements made for regulatory controls be as small as possible.
Alternatives proposed but also limited
Several alternatives have been proposed/used but they are also limited and complex to implement.
The Size-Specific dose estimate (SSDE), which aims to give an adjusted CTDI to the patient’s morphology, but has the same limitations as the CTDI and does not take into account variations in scan length, modulation and changes in the patient’s morphology in the Z axis.
The effective dose, which can be useful to compare modalities and different radiological and/or radioactive risks but cannot be used to estimate individual risks and is misleading on relatively high doses delivered locally to some organs despite a low overall effective dose of the examination. In addition, although this method is convenient and fast, calculation of effective dose directly from the DLP should be considered with caution.
The use of dosimetry on anthropomorphic phantoms can be a solution but the variability of morphology and organ size from one patient to another limits its use.
Monte Carlo simulation, which aims to evaluate the doses delivered in CT, has the advantage of being customizable, for example by using patient images directly, but data concerning the scanner’s x-ray spectrum, its filtration and the beam geometry are necessary and are not always provided by the manufacturers. Moreover, this method requires a certain level of expertise and is time consuming.
In vivo dosimetry with direct skin dose measurements can give a good estimate of the dose to superficial organs such as the lens of the eye, thyroid gland, breast and male gonads.
Real-Time Personalized CT Dosimetry: The Way Forward
Patient dose estimates must be:
– Equipment specific, taking into account parameters such as filtration and beam spectrum and geometry
– Patient-specific, using individual models created from patient cross-sectional data
– Protocol specific, taking into account all protocol parameters
– Available in real time and displayed immediately after the examination Intuitive software directly integrated into the scanner control stations is required to provide organ dose results based on personalized dosimetry with patient imaging data including the help of artificial intelligence.
Real-time personalized dosimetry taking all these parameters into account and able to provide accurate organ dose estimates may lead to a paradigm shift in CT patient dosimetry.
Read the article: https://pubmed.ncbi.nlm.nih.gov/32932380/
Fibermetrix is developing solutions to provide reliable real-time dosimetry information in CT.