By Dr Entesar Zawam Dalah, Assistant Professor, Medical Diagnostic Imaging Department, College of Health and Science, University of Sharjah, UAE

Unlike conventional radiographic images, medical images acquired with computer tomography (CT) never look over- or under-exposed i.e., never too dark or too bright. This is due to the fact that CT data is normalised to represent a fixed amount of attenuation relative to that of water. Normalisation ensures that CT images always appear properly exposed despite the amount of radiation exposure (dose).

Thus, relying on the visual appearance only of a CT image one can never determine if a patient is being exposed to excessive radiation. Due to this, CT dose reduction strategies, quality control (QC) tests, and quality assurance (QA) tests for patients undergoing CT examination became a central worldwide focus.

While CT is indispensable in the medical diagnostic field, numerous articles have raised attention to the significance of ionising radiation dose that patients receive while undergoing CT examinations. Recently, couple of epidemiology studies have reported an increasing risk of leukemia and brain cancer in children, adolescents, and young people with the increasing dose and use of CT scans.

One of the follow-up cohort studies involved 180,000 young people who underwent CT scan in the United Kingdom during the period of 1985-2002. This study reported an increased risk of leukemia and brain cancer with increasing doses of radiation from exposure to previous CT scans.

The other study by Mathews et al reported that the increased incidence of cancer in their cohort study was mostly due to CT scan exposure. This cohort study included 10.8 million children and adolescents in Australia, collected in the period between 1985 to 2005.

The notion of ionising radiation-inducing cancer is well known, which in part results from the disrepair of the DNA damage caused by the hydroxyl radicals, which results from radiation interactions with the cellular water molecules, known as the indirect interaction. Alternatively, radiation can interact directly with the DNA and lead to even more severe damages in a shorter period of time. Most of radiation-induced DNA damages are rapidly repaired by various systems within the cell, but DNA double-strand breaks are less easily repaired and can lead to the induction of point mutations, chromosomal translocations, and gene fusion.

It is also noticeable that CT usage is in drastic increase worldwide, counting as high as 10% to 15% per year. Further, CT technology is rapidly evolving, leading to the explosion of new clinical applications including cardiac CT, and multiphase CT exams with and without contrast. Moreover, radiation dose levels imparted in CT has been shown to exceed those from conventional radiography and fluoroscopy - with as low as 0.2 mSv for a typical chest x-ray and as high as 11 and 25 mSv for abdominal and pelvic CT, and CT angiography, respectively.

It is also striking to learn that radiation doses from identical CT procedures can vary by a factor of 10 from one facility to another and even within a facility. All of the above have raised a compelling need to understand, optimise, and document patient CT dose. The other factors that are usually overlooked are issues related to quality control, lack of training, and the overprescribing of CT imaging. As such, radiologists and clinical health providers need to be aware of, and able to justify, the doses delivered during various types of CT studies performed at their institutions, particularly when it concerns paediatric and adolescent patients.

Nowadays modern CT scanners are equipped with dose data display in the form of: (a) volume computed tomography dose index (CTDIvol), and (b) dose length product (DLP). While the CTDIvol dose index is being referred to and displayed in mGy, the second dose product is displayed in (mGy.cm) accounting for the scan length product in centimetres. Both the dose indices displayed on the CT console are not demonstrating patient dose. Actually, CTDIvol only establishes CT radiation output that is measured in a reference and standardised way independent of organ/tissue radio-sensitivity, and patient-specific size.

The CTDIvol and DLP dose indices displayed on the CT console are for a polymethyl methacrylate (PMMA) cylinder reference phantom, often called head (16 cm) and body (32 cm). Because both CT dose indices are independent of patient size, which is a vital parameter for estimating patient CT absorbed dose, interpreting CTDIvol for smaller size patients or paediatric patients could lead to under estimation of radiation dose by a factor of 2 – 3 using the PMMA reference phantom.

This being said, CTDIvol and DLP are sensitive to changes in scan parameters such as kVp, mAs, gantry rotation time, pitch, and bowtie filter. In light of this, the patient absorbed and effective dose in mGy and mSv, respectively, has to be determined for individuals undergoing CT studies taking into consideration the institutional technical protocol of each and every CT study, patient size, and body scanning area including head, neck, chest, trunk, abdomen and pelvic.

It is essential to remember that those pretty CT images are not needed for all diagnostic tasks, but rather, the choice should be made between low noise and low dose, depending on the diagnostic task. Setting the parameter for “increased” image quality (i.e., lower noise) will result in more dose, and vice versa. So, when it comes to a paediatric CT exam, the procedure must be justified and protocols must be implemented to reduce the dose for the same image quality as in adults. We need to work toward a safer environment for our patients by keeping CT radiation doses as low as reasonably achievable (ALARA).
References available on request.

Dr Entesar Zawam Dalah is a Speaker at the Radiology Safety Conference held as part of Patient Safety Exhibition in Dubai, on 26th October, 2017.