Chest CT features of the novel coronavirus disease (COVID-19)

A new type of coronavirus (2019-nCoV) is rapidly spreading worldwide and causes pneumonia, respiratory distress, thromboembolic events, and death. Chest computed tomography (CT) plays an essential role in the diagnosis of viral pneumonia, monitoring disease progression, determination of disease severity, and evaluating therapeutic efficacy. Chest CT can show important clues of 2019-nCoV disease (also known as COVID-19) in patients with an appropriate clinic. Prompt diagnosis of COVID-19 is essential to prevent disease transmission and provides close clinical observation of patients with clinically severe disease. Therefore, radiologists and clinicians should be familiar with the CT imaging findings of COVID-19 pneumonia. Herein, we aimed to review the imaging findings of COVID-19 pneumonia and examine the critical points to be considered for imaging in cases with COVID-19 suspicion.

in the early phases of the COVID-19 infection or in the presence of disease with a low viral load, the RT-PCR test may be negative, but CT may reveal significant findings [14,15]. However, in patients with COVID-19, especially in the early stage of the disease, chest CT can be found entirely normal, and normal CT does not rule out the disease. Moreover, chest radiography can be used in COVID-19 patients, but its diagnostic value is very limited. Therefore, it is recommended to use chest radiography in the assessment of lung involvement in paediatric and pregnant patients or in the follow-up of hospitalized patients, especially in case of severe or critical illness [16].
With growing global concerns about COVID-19 pandemic, radiologists and clinicians should be familiar with the imaging findings of COVID-19 pneumonia. Herein, we aimed to review the imaging findings of COVID-19 pneumonia and examine the critical points to be considered for imaging in cases with COVID-19 suspicion.

Usual imaging findings
The reported prevalence of the common CT findings is shown in Table 1.  Zhao et al. [24]  lung parenchyma because of partial filling of air spaces, interstitial lung disease, increased capillary blood volume, partial alveolar collapse, or a combination of these, the common factor being the partial displacement of air [17]. Ground glass opacities with a peripheral and multi-lobar distribution is a common finding in COVID-19 patients, and GGO is usually more prominent in lower-middle lobes and posterior lung areas (Figures 1a and 1b). The reported prevalence of GGO varies between 46% and 100%, and GGO has usually seen in the early phases of the disease and/or mild pulmonary infection [9][10][11]13,14,. As the disease stage and severity increase, usually consolidation and/or interstitial thickening (crazy paving pattern) begins to appear within the GGO areas. Besides, while GGO or consolidation are healing, it can be seen as a GGO containing fibrotic band or atelectasis in the lung [31].

Consolidation
Pulmonary consolidation is defined as an increase in lung density with obscuring vascular margins in the lung parenchyma due to the complete replacement of air spaces (alveolus and alveolar sacs) by fluid, pus, blood, or cells [17]. Peripheral, segmental, or sub-segmental consolidation areas with the multi-lobar distribution are usually seen in cases with COVID-19 pneumonia, especially in older patients and patients with severe or critical disease ( Figure 2) [11,[14][15][16]31,39]. Consolidation areas are usually seen in mixed patterns with GGO, and the new emergence of consolidation areas in the followup imaging of COVID-19 patients was considered as an indicator of progressive disease [11]. Besides, consolidative lesions have been reported in COVID-19 patients with a longer time interval between onset of symptoms and chest CT scan, or patients over 50 years of age [31,39]. The prevalence of thromboembolic events in patients with COVID-19 has been shown to increase compared to the healthy population. Therefore, when CT shows peripherally located, triangular-shaped consolidation areas, pulmonary thromboembolism, and infarction (Hampton hump) should be considered in the differential diagnosis [44,45].

Air bronchogram and airway changes
Air bronchogram is the appearance of air-filled and low attenuated bronchi on a background of the high attenuated opaque lung, and the air bronchogram has almost always been caused by airspace pathology [17]. Air bronchogram sign has been frequently reported in cases with COVID-19 Figure 1. a) A 48-year-old man with COVID-19 presented with cough, fever, myalgia, and malaise for 3 days. Axial chest CT image shows multifocal, peripheral-peribronchovascular ground-glass opacities (red frames). b) The magnified image of the right lower lobe shows ground-glass opacity superimposed with interlobular septal thickening and prominent intralobular lines (red arrowheads), which indicates a crazy-paving pattern.

Pulmonary vascular enlargement
Pulmonary vascular enlargement (PVE), also known as pulmonary vascular widening, is defined as the widening of the subsegmental pulmonary vessel (pulmonary arteries and veins) with a diameter of more than 3 mm around and/or inside the opacity on CT (Figures 4a and 4b) [24,25,27,32,35,39]. Although the pathophysiology of PVE is not fully understood, PVE plays a potential diagnostic role for COVID-19. Bai and colleagues [32] had compared the CT findings in patients with COVID-19 pneumonia and non-COVID-19 pneumonia, and they found that the PVE was significantly associated with COVID-19 (P < 0.001, 59% vs. 22%

Pulmonary embolism
Life-threatening thromboembolic events are increasingly identified in COVID-19 patients [44,45,47]. Recently, Grillet and colleagues [44] reported the prevalence of pulmonary embolism (PE) as 23% (n = 23/100) in COVID-19 patients with severe clinical features, and they reported that COVID-19 patients with PE, had a higher requirement for mechanical ventilation than those  . Axial chest CT image shows peripheral and peribronchovascular distributed focal ground-glass opacities in both lungs, and focal vascular enlargement is seen inside the ground-glass opacity in the right lower lobe (a, red frame). The magnified image of the red frame shows vascular enlargement inside the ground-glass opacity (b, arrows  [44,45]. Therefore, pulmonary CT angiography is essential for evaluating lung parenchyma and vascular complications for COVID-19 cases with severe and/or critical clinical features and high D-dimer levels (Figures 6a, 6b, and 6c). Recently, the British Thoracic Imaging Association has suggested that if pulmonary CT angiography (CTA) is required in a potential COVID-19 patient, prior to CTA, an unenhanced chest CT should be performed because mosaic attenuation in CTA can be confused with GGO [48].

Unusual imaging findings
The reported prevalence of the unusual CT findings is shown in Table 2.

Pleural thickening and effusion
Pleural pathologies such as pleural effusion and focal pleural thickening have been rarely reported in cases with COVID-19, and pleural pathologies are usually seen in the later stages of the disease [21,31,39]. The presence of pleural effusion is thought to be a sign of poor prognosis in COVID-19 pneumonia, and the prevalence of pleural effusion in COVID-19 patients has been reported to range from 0% to 20% [11,13,14, (Table 2). Besides, the pleural effusion and thickening might be associated with dense pleural inflammation in patients with COVID-19 (Figures 6a, 6b, 6c, 7a, 7b) [24].

Air bubble sign
The air bubble sign has been defined as a small aircontaining space that may be a pathological enlargement of bronchioles inside the opacity [25,28]. Air bubble sign is named in the literature under different names (cystic air space, round cystic changes, air-containing space, sievehole sign, and vacuolar sign). Although its pathophysiology is unclear, it may be related to the pathological expansion of alveolar sacs or bronchioles or the absorption process of consolidation, and the reported prevalence of air bubble sign in patients with COVID-19 pneumonia ranges from 8.1% to 54.8% (Figure 8) [25,28,33,34,35,39] (Table 2).

Nodules
A nodule refers to opacity in the lung parenchyma, less than or equal to 3 cm in diameter [17]. Shi and colleagues [34] reported that lung nodules were detected in 6% of patients with COVID-19 pneumonia. Similarly, Li and colleagues [30] reported that pulmonary nodules with or without visible halo sign were detected in 7.2% of patients with COVID-19 pneumonia (Figure 9). The halo sign indicates GGO surrounding a pulmonary nodule or mass and usually seen in patients with invasive fungal infections, hypervascular pulmonary metastases, and organizing pneumonia [17]. Although it is reported in the literature that tree-in-bud pattern nodules can be seen in patients with COVID-19 [24,28,33,35], superimposed bacterial infections or aspiration should be considered first when these nodules are detected.

Reversed halo sign
The reversed halo sign (RHS) indicates a central GGO surrounded by denser ring-like (crescentic shape) consolidation and also known as the Atoll sign [17]. Although RHS is usually seen in patients with cryptogenic organized pneumonia (COP), it has been reported in several cases with COVID-19 [13,20,27,32,35]. While Bernheim et al. [13] reported that the reversed halo sign was detected in 2 of 121 patients (1.7%) with COVID-19 pneumonia, Wang et al. [35] reported the prevalence of reversed halo sign as 15.1% (14 of 93 patients) ( Table 2) ( Figure 10).

Spider web sign
The "spider web sign" was defined by Wu and colleagues [14], and they described it as an angular or triangularshaped peripheral GGO with interstitial thickening like a spider web in a corner, but its pathophysiology has not been clearly revealed. They reported that the reversed halo sign was detected in 20 of 80 patients (25%) with COVID-19 pneumonia [14]. Moreover, Li et al. [30] detected the spider web sign in 21 of 83 (25.3%) patients with COVID-19 pneumonia. Apart from these 2 studies, there are no other studies in the literature describing or investigating the "spider web sign". Therefore, this finding Figure 6. A 40-year-old female patient was presented with complaints of fever that had been ongoing for 5 days, and newly emerged severe dyspnoea. The patient's D-dimer level was found to be very high and contrast-enhanced pulmonary CT angiography was obtained. a) Axial CT image at the level of the origin of the middle lobe bronchus with mediastinum window settings shows bilateral diffuse pulmonary embolism (red arrows), bilateral mild pleural effusion, and subpleural triangularshaped opacity with a reverse halo sign in the left lung compatible with pulmonary infarction (red frame). b, c) Axial CT images of the same patient with lung window settings show multiple subpleural consolidation and ground-glass opacity areas (red frames). The nasopharyngeal swab test of the patient was positive for 2019-nCoV.
needs to be investigated in patients with non-COVID-19 pneumonia and COVID-19 pneumonia.

Lymphadenopathy
Mediastinal lymphadenopathy has been infrequently reported in COVID-19 patients, and lymphadenopathy has been shown as a risk factor for severe and progressive COVID-19 pneumonia [14,31,37]. In the literature, the frequency of mediastinal lymphadenopathy has been reported between 0% and 29% in patients with COVID-19 [11,13,14,. Lymphadenopathy is a rare finding in viral pneumonia, and the detection of pleural effusion, lung nodules with a tree-in-bud pattern, and ill-defined micronodules may suggest bacterial superinfection [31][32][33][34].

Pericardial effusion
Pericardial effusion has been rarely reported in COVID-19 patients, which may indicate the occurrence of myocardial and/or pericardial inflammation [14,18,23,30]. Li et al. [30] have reported that COVID-19 patients with severe and critical diseases showed a higher frequency of pericardial effusion than noncritical patients. Recently, Xu et al. [23] found the pericardial effusion in 1 of 90 (1.1%)    . Axial chest CT image shows mixed opacity (ground-glass opacity with consolidation) area in the right middle lobe (red frame), and that includes cystic air space (air bubble, red arrow).
patients with COVID-19 pneumonia. However, elderly patients, and especially those with additional heart failure, may have pericardial effusion independent of COVID-19.

Structured CT reporting for assessment of COVID-19
Although chest CT findings in COVID-19 patients can partially overlap with other diseases, especially other types of viral infections, COVID-19 may have characteristic CT features that are less common in other diseases [32]. Various recommendations have been published to standardize the structured CT reporting for the suspected COVID-19. The first of these is the Radiological Society of North America (RSNA) expert consensus statement, which suggests a standardized reporting and a classification of imaging features for COVID-19 pneumonia (i.e. negative for typical appearance, vague appearance, atypical appearance, and pneumonia) [49]. Salehi and colleagues [50] have  (Table 3). However, the groups in this classification are based on the results obtained from a systematic review, and clinical success and applicability of this classification are unknown. Recently, the Dutch Radiological Society defined a new classification (CO-RADS) for pulmonary involvement in cases presenting with moderate to severe symptoms of COVID-19 [51]. In this standardized assessment (CO-RADS), a substantial agreement was found among 8 observers (Fleiss' kappa of 0.47, 95% CI 0.45-0.49) and the discriminatory power of the CO-RADS for diagnosing COVID-19 was high (with an area under the curve value of 0.91, 95% CI 0.85-0.97). Therefore, the authors stated that CO-RADS is very suitable for use in clinical practice [51] (Table 4).

The severity of pulmonary involvement on CT
The visual (semiquantitative) evaluation of disease severity on chest CT can reflect the clinical classification (mild, common, severe, or critical disease) and prognosis of patients with COVID-19 pneumonia [11,30,41]. Therefore, the staging of lung involvement has been defined in COVID-19 patients. In these staging methods, the percentage of each lung lobe involvement is assessed, Figure 9. A 36-year-old woman presented with complaints of cough and fever for 2 days. Axial chest CT image shows a nodule with halo sign in the right upper lobe (red arrow). The nasopharyngeal swab test of the patient was positive for 2019-nCoV, and the nodule disappeared at follow-up. and the portion of the total lung involvement is calculated semiquantitatively (visually). For example, Chung et al. [11] and Li et al. [41] used the same scoring as follows: none (0%); a score of 0, minimal (1%-25%); a score of 1, mild (26%-50%); a score of 2, moderate (51%-75%); a score of 3, and severe (76%-100%); a score of 4. An overall lung involvement score was reached by summing the 5 lobe scores (0-20). Besides, quantitative CT assessment methods and artificial intelligence applications can be successfully applied in the diagnosis of COVID-19 and in the evaluation of disease severity [37,52,53].

Radiation concern
Exposure to medical radiation is a concern in patients with suspected or diagnosed COVID-19 pneumonia, especially in infants, children, young adults, and pregnant patients. Therefore, it is necessary to prevent or reduce radiation exposure [6,7,20,54]. First, it should be distinguished appropriately in which patients need imaging and in which patient imaging is not needed. In cases with COVID-19 suspicion, imaging methods should be avoided if clinical follow-up and management will not change. In a multinational recommendation that is newly defined and widely used by Fleischner Society, imaging is not recommended for cases with suspected of COVID-19 without risk factors (such as diabetes, older age, chronic obstructive pulmonary disease) and mild clinical findings. Moreover, imaging methods are indicated in a COVID-19 patient whose respiratory condition has worsened [6]. Second, chest CT should be obtained as low as possible radiation dose in cases with CT indication. Dangis et al. [54] showed that chest CT with low-dose settings would provide the diagnosis of COVID-19 pneumonia with similar accuracy as standard-dose chest CT.

Contamination of radiology units
Radiography machines and CT scanners are frequently used in patients with suspected COVID-19 during the pandemic. Therefore, vital precautions should be provided for COVID-19 in radiology units [6,55]. First, ensuring the health of radiology workers is essential for the best care of patients. For this reason, staff working in radiology units should use necessary personal protective equipment according to their contact with the patient. Besides, all patients with a suspicion for COVID-19 who presented to the radiology unit should use a standard surgical mask. Second, contamination of radiography machines and CT scanners are a major concern. Therefore, it is necessary to use a dedicated radiography machine and CT device in patients with suspicion for COVID-19 to reduce the risk of disease transmission. Also, these dedicated devices should never be used in patients without suspicion for COVID-19. Third, to reduce the risk of contamination, the CT scanners and radiography devices and rooms must be deeply cleaned and disinfected according to standard protocols after each exam. If patient masking is available, active (mechanically) air exchange procedures are not required [6,55]. After each imaging procedure, the room downtime for the decontamination of room and devices is approximately 30 min.

Conclusion
In conclusion, early diagnosis of COVID-19 is essential to prevent transmission and provide close clinical observation of the patients. While the gold standard for the diagnosis of COVID-19 is RT-PCR, CT should be obtained in a limited population (in COVID-19 patients who have worsening respiratory symptoms, in patients with moderate-severe clinical features of COVID-19, in medical triage of COVID-19 suspected cases with moderate-severe clinical features, and a high pretest probability of COVID-19 in the lack of personal protective equipment or availability of RT-PCR test) due to radiation concern. Although bilateral and peripheral GGO without or with superimposed interstitial thickening (crazy-paving pattern), and/or consolidation with the multi-lobar distribution is the common CT features in COVID-19, chest CT features can vary among patients. Moreover, chest CT is not just a diagnostic or screening method in cases with suspected COVID-19. In patients with COVID-19, chest CT is especially crucial in determining the severity of pneumonia, which has a prognostic impact [11,37,41].