JTI Blog

Current events in cardiopulmonary radiology, updates about the journal’s web site features, and links to other web sites of interest to cardiopulmonary radiologists.

Sunday, June 8, 2014

Radiation: Rationem Practicam, A New JTI Blog Series by Mannu Kalra
II: Personalizing Chest CT Protocols

It is not my intent to enter into the minefield of risks of low radiation doses such as those associated with CT scanning. Yet I must re-affirm my belief that prudent use of radiation dose in CT scanning must not be an optional aspect of any CT practice. On the other hand, while quest to bring CT doses for certain thoracic indications closer to a few chest radiographs has already begun, until that point of “miniscule” dose is reached or crossed, a certain caution is important to protect the diagnostic information from this precious imaging modality. This aspect of prudent radiation dose utilization brings us to an important matter of personalizing CT protocols depending on clinical requirement or need. Stratification of CT protocols can and does help in ensuring “just-enough” doses from CT. I like to qualify the ALARA as radiation dose as low as reasonably achievable for specific clinical indication.

Before the scanner description and scanning parameters populate a CT protocol, comes the clinical indication or reason for scanning. No other scanning attribute is as important as a justifiable clinical indication for CT scanning. The best clinical justification for CT scanning results in the best benefit to risk ratio for CT. Based on the similar clinical indications, we should stratify the chest CT protocols so as to enable maximum efficiency and least errors in selection of appropriate CT radiation doses.

Several different clinical indication-based CT protocols can be applied to chest imaging so that we do not use a “one size fits all” approach. The thoracic imaging division should at least have a few specific CT protocols such as routine chest, CT pulmonary angiography, lung nodule follow up CT, lung cancer screening CT, diffuse lung disease CT (also called high resolution CT of the chest), tracheal protocol CT, and chest wall protocol. Such stratification of protocols can help optimize radiation dose according to clinical requirements.

For example, radiation dose for both lung nodule follow up CT and lung cancer screening CT should be lower than the routine chest CT protocol. Creation and review of maximum intensity projection (MIP) images helps improves detection of lung nodules while averaging out image noise in low dose images. Likewise, a distinct CT pulmonary angiography protocol must utilize weight based or software based lower tube potential (kilovoltage) as well as limited scan coverage up to lung bases only as compared to routine chest CT protocol. Thinner sections help reduce partial volume artifacts and improve sensitivity for detection of pulmonary embolism.

CT for diffuse lung disease is another distinct protocol that should spell out policies for acquiring expiration phase images either as a step-and-shoot (non-helical or axial) acquisition of representative sections with gaps or as a low-dose volumetric (helical) acquisition from lung apices to bases. To maximize spatial resolution, a diffuse lung disease protocol should also involve reconstructing thin sections with higher spatial frequency kernels (aka sharper algorithm, kernel or filter) at smaller reconstructed field of view from rib-to-rib coverage. Prone images likewise may not be necessary in all patients.

Evaluation of the trachea with CT, though not as frequent as some of the other indications, nonetheless warrants an additional protocol to ensure optimum scan coverage (from base of the neck to the lung bases in inspiration and up to the bifurcation in dynamic expiration). The importance of appropriate breathing instructions and compliance cannot be over-emphasized for chest CT, particularly for diffuse lung disease and tracheal protocol CT examinations.

Once CT protocols are stratified according to clinical indications, we should then sub-stratify these individual protocols according to the scanner types (number of sections (4-, 8-, 16-, or 64-row multidetector row CT) and vendors. Different CT systems have unique capabilities and therefore unique requirements for scanning parameters. Thus, for example, lung nodule follow up protocols should be created for different CT systems (separate protocols for Siemens 64-section versus Toshiba 64-section). This may sound onerous and forbidding but in reality is not an insurmountable task.

In the next blog, I will discuss the elements of good scan protocols, including the need to follow some uniform protocol codes that can simplify our lives in terms of protocol and dose comparisons.

Email me at MKALRA@mgh.harvard.edu with your comments and questions.