Chest HRCT is based on the contrast between the low attenuation of air and the relatively higher attenuation of vessels, airways and interstitial structures. Usually, lung CT examinations are performed in suspended end-inspiration to maximize the natural contrast between air and pulmonary structures (Fig. 1); CT acquisitions at suspended end-expiration can be additionally required to detect diseases of both large and small airways.

Inadequate respiratory maneuvers may increase lung attenuation - simulating disease in normal patients (Fig. 2); in addition, inadequate respiratory maneuvers - due to reduced lung inspiration - can decrease pulmonary attenuation, leading to normal contrast in patients with disease; finally, inadequate respiratory maneuvers - due to anxiety or restlessness of patients - can reduce the quality of images for the presence of motion artifacts (Fig. 3) [3]. Therefore, patients have to be instructed before the acquisition and coached during it (Fig. 4) [3].

Principal mistakes and artifact HRCT are summarized in Fig. 5.

False Mosaic Attenuation

It may be related to air trapping (caused by airways obstructive diseases) or may be associated with infiltrative diseases or vascular conditions. Inadequate inspiration, revealed by a partially collapsed posterior tracheal wall, may reproduce a false mosaic attenuation (Fig. 6). In some interstitial diseases, ground-glass opacities reflect intralobular thickening due to fibrotic abnormalities; a false ground-glass may be related to inadequate inspiration (Fig. 7).

Supine/Prone position: the effect of lung "dependent density"

HRCT scans obtained with the patient in supine position are generally enough for diffuse lung disease diagnosis; additional prone scans could be necessary to diagnose or exclude subtle disease in the posterior part of the lung. Atelectasis is commonly seen, in the dorsal pulmonary regions of both healthy and diseased subjects, like a “dependent density” or “subpleural line”. These normal findings are related to gravity and may mimic the appearance of early lung fibrosis, such as asbestosis. In the HRCT scan obtained with the patient prone, the normal dependent density disappears and can be differentiated from the true disease [4-5-6].

In some cases, a prone acquisition may use to detect underlying honeycomb or reticulations; more in detail, dependent regions of the lung may produce along the dorsal regions increased density or ground-glass opacification - hiding true fibrotic alterations (Fig. 8).

In addition, these areas of increased attenuation – related to dependent lung regions – are observed only along the dorsal regions, and are not recognizable in the ventral parts. During a chest HRCT examination, the patient's conditions need to be carefully evaluated.

Body size

Image noise, or quantum mottle, is the most common imaging artifact encountered in the bariatric population; it results from an insufficient number of photons reaching the detector (Fig. 9). The noise can be reduced by increasing the tube current (“mA”), the time per rotation (“s”) and the tube voltage (“kV”). Similar to increasing the time per rotation, lowering the pitch results in less-noisy images at the expense of increased scanning time and increased radiation dose to the patient.

For larger patients, excess soft tissues may fall outside the scan FOV, but the scanner reconstruction algorithm will assume that all of the attenuations occurred within the scan FOV. As a result, the periphery of the reconstructed image will appear to have a substantially higher attenuation, generating a truncation or cropping artifact. It’s important to be aware that bariatric patients often present some related factors that can affect the examination accuracy, like cardiac output, intravascular volume and total mass of iodine delivered to the target organ [7].

False measurements

During follow-up CT for monitoring nodules evolution, false measurements may be caused by the use of different collimation or kernel. An important recommendation is to follow Fleischner 2017 guidelines (Fig. 10), in order to properly measure the diameters. In oncologic CT imaging, diameters volume, and appearance of tumors and nodules do not depend only on section thickness and CT acquisition parameters; they also vary with patient’s respiratory maneuvers [8]. 

In cases of pulmonary embolism, CT protocol optimization plays a crucial role in detecting vascular filling defects. Most common artifacts are:

- respiratory motion artifacts: in the craniocaudal acquisition, they affect mainly the lower lobes because the patient has more difficulty holding breath [9].

These artifacts are seen with lung window settings and can create the “seagull” sign (Fig. 3b); also low-attenuation abnormality due to partial volume averaging of vessel and lung can simulate embolism [8].

- streak artifacts: in craniocaudal acquisitions, they occur at the level of the superior vena cava with a highly concentrated material and a fast rate of injection. Reduced iodine concentration and caudo-cephalic scanning can decrease such artifacts.

- inadequate delay of injection: with a craniocaudal acquisition, too short a delay may lead to insufficient opacification of the upper pulmonary arteries and too long a delay will preclude correct opacification of the lower pulmonary arteries. 

- factors intrinsic to the patient; a functional venous systemic compression at the thoraco-brachial junction due to the patient's arms above his or her head, any organic cause of obstruction of the systemic veins (as in superior vena cava syndrome) and an increased pulmonary arterial pressure with a right-to-left shunt through a patent foramen ovale [10], can cause delayed enhancement of the pulmonary arteries.

- stair step artifact, low-attenuation lines seen traversing a vessel on coronal and sagittal reformatted images, that can be eliminated or reduced by reconstructing the raw data with a 50% overlap prior to 3D reconstruction [10].