Air bubble artifact
Citation, DOI, disclosures and article data
At the time the article was created Daniel J Bell had no recorded disclosures.View Daniel J Bell's current disclosures
The air bubble artifact on CT is due to the presence of abnormal gas in the oil coolant which surrounds the x-ray tube. The artifact manifests as subtle low density, which has only been described on brain scans.
The x-ray tube in a CT scanner is prevented from overheating by a heat exchange system which uses oil as its coolant. The abnormal bubbles of air/gas in the system subtly modify the transmission of the primary x-ray beam, decreasing its attenuation by up to 3 HU. The number and precise location of the bubbles may vary over time - due to their movement in the coolant - so that fluctuating attenuation of the x-rays occurs as coolant circulates and the tube rotates. Therefore the position and severity of the artifact also varies. As the effect on the attenuation of the x-ray beam is very mild this artifact has only been seen when narrow window widths are used, which for practical purposes is solely CT brain studies, primarily on "stroke" window settings.
Formation of air bubbles
Gas/air bubbles can arise within the coolant oil via several different mechanisms 3:
- during CT service/repair, e.g. oil changes/top-ups
- loss of integrity of the tube envelope/heat exchanger: it is designed to be a self-contained unit with no communication with the external environment
- spontaneous formation of gas within the oil when in situ, due to vaporization in the system, increasingly common as the tube ages
Detecting the artifact
It can be difficult for radiologists/radiographers to realize that this artifact is present, as true subtle low density on CT brain is often seen in early ischemia and other pathologies. Radiologists or clinical referrers may become suspicious when consistent subtle findings on CT are not visible on repeat CT or MRI.
A study from 2017 found that all scans with confirmed artifact had a characteristic stripe-like pattern of hypodensities, passing in an anteroposterior orientation, on the sagittal reformatted images. This stripe is not seen when the low densities of ischemia are reviewed on the reformatted sagittal dataset.
If there is suspicion that the artifact might be present then the vendor needs to be contacted for testing of the scanner by qualified personnel. A service engineer is able to perform a 'gas detection test' to determine how many gas bubbles are in the coolant.
Newer scanners are now available which are able to give a readout of the number of gas bubbles in the oil coolant, and can alert the scanner vendor if the quantity of gas reaches a predetermined critical threshold.
Removing the artifact
Resolving the artifact requires an engineer to replace the oil and treat any underlying defect in the system e.g. a leak in the tube housing.
History and etymology
This artifact was first detected in 2015, and described in 2016 by two separate groups (one in Scotland, the other in the US) in two different case-series, each seemingly unaware of the other 1,2. It has now been reported by multiple centers. It remains unknown why this artifact was not reported previously, even allowing for its subtle nature.
- 1. Hedrick WR, Markovic MA, Short JH, Vera CD. Computed Tomography Artifact Created by Air in the X-ray Tube Oil. (2016) Journal of computer assisted tomography. 40 (1): 67-70. doi:10.1097/RCT.0000000000000315 - Pubmed
- 2. McVey S, Kanodia AK, Dundas S, et al. "Air bubble artefact": a new type of artefact on CT head examination. (2016) Clinical radiology. 71 (10): 1059-65. doi:10.1016/j.crad.2016.06.118 - Pubmed
- 3. Trieu N, Xia R, Loneragan R, et al. Artifact on CT brain images caused by the presence of air bubbles in the cooling oil of the X-ray tube. (2017) Journal of medical imaging and radiation oncology. 61 (2): 197-203. doi:10.1111/1754-9485.12545 - Pubmed
- 4. Wong SSM, Cheung JSW, Lee TWK, et al. CT brain artifact due to air bubbles in the oil cooling system: characteristic band-like configuration on sagittal reformatted image. (2018) Japanese journal of radiology. 36 (2): 90-95. doi:10.1007/s11604-017-0702-3 - Pubmed