Coronavirus Disease 2019 (COVID-19) Radiologic Images

Our knowledge of the imaging characteristics of COVID-19 and the way we should incorporate imaging in the management of COVID-19 patients is rapidly evolving since the first case was diagnosed in late December 2019. The Centers for Disease Control and Prevention (CDC) does not recommend chest CT or chest x-ray to diagnose COVID-19 and emphasizes that the only specific method to diagnose the disease is viral testing, [1]  and the Fleischner Society consensus statement published on April 7, 2020, recommends against the use of imaging in patients with suspected COVID-19 and mild symptoms. However, imaging can be used as a diagnostic modality in the following situations: (1) when there is concern for disease progression, (2) when there is worsening respiratory status in known COVID-19 cases, and (3) in resource-constrained areas where imaging can be used for triage of individuals with suspected COVID-19 based on a high pretest probability and moderate to severe symptoms.[2]  (See the CT images in the Chest CT Features section, below.)

There are growing publications on the validity of chest CT and chest x-ray on detection of COVID-19 patients. In a series of 51 patients in whom chest CT and reverse transcription-polymerase chain reaction (RT-PCR) assay were performed within 3 days, CT showed a sensitivity of 98% for COVID-19 detection.[3]  A review of 1014 chest CTs from Wuhan, China, obtained for assessment of COVID-19 echoed similar findings, with a 97% sensitivity in detection of COVID-19.[4]  In another study, chest x-ray showed a 69% sensitivity for detection of COVID-19. It is very important to mention that imaging features of COVID-19 are not 100% specific.[5] Most of the same imaging characteristics can be seen in other types of pneumonia, such as viral pneumonia from influenza. However, there are some characteristic features that can help physicians raise high concern for COVID-19 (see the images above). From a public health perspective, prompt identification of COVID-19 cases and rapid isolation are crucial in decreasing the rate of spread of this highly communicable disease, highlighting the importance of familiarity with the imaging features of COVID-19.

Physicians need to be aware that chest x-ray is insensitive in detection of positive COVID-19 cases in the early phase of the disease.[5]  Although CT is a more sensitive modality than chest x-ray for detection of COVID-19, the final decision on the choice of imaging modality should be based on several factors and scenarios at the point-of-care. For example, in areas with decent cooperation of the population with social distancing where patients are asked to seek medical care only when they have advanced symptoms (such as New York or Italy), chest x-ray has often shown abnormal findings and has been helpful in detection of positive cases. Chest x-ray is also helpful in inpatient settings for assessment of the progression of disease or possible superimposed complications such as bacterial infection, pleural effusion, or pneumothorax.

In a setting where the goal is to isolate all cases in the early phase of the disease and where reliability of viral testing is limited (such as Wuhan, China), chest CT can be considered a superior modality given its ability for better detection of early-phase pneumonia. There is no doubt that several other factors play significant roles in final decision making regarding the imaging modality of choice for management of COVID-19 patients, such as resource availability, image interpretation expertise in the region, and the importance of equipment portability based on patients’ clinical condition and ease of transportation. 

The most common CT features reported in the currently available literature are ground-glass opacity (GGO, 20-62%), mixed GGO and consolidation (up to 42%), and consolidation (up to 50%).[4, 6, 7, 8]  CT findings are evolving during the course of disease, with the most lung involvement being seen during illness days 6-11.[6]  While GGO can be seen throughout the process of disease, consolidations are more common after day 5. Interlobular thickening is more common 6-17 days after development of symptoms. More bandlike opacities are expected in the late phase of the disease.[6]  Some other less common CT textures that can be seen in COVID-19 include crazy-paving pattern, reversed halo (atoll) sign, and bronchial wall thickening [7] . 

Some CT features are not common in COVID-19 and, when seen, can raise concern for alternative diagnosis or superimposed bacterial infection. For example, pleural effusion has been reported in 7% of COVID-19 cases. Pneumothorax, mediastinal adenopathy, cavitation, discrete nodules, and lobar pneumonia are also not expected in COVID-19 chest CT scans.[6, 7]  CT findings are usually bilateral and multifocal (50-75%) with peripheral/subpleural distribution.[9]  There is a tendency for involvement of the lower lobes in COVID-19.[7] . 

(See the CT images below showing characteristic findings of COVID-19.)

Axial chest CT demonstrates patchy ground-glass opacities with peripheral distribution.

Coronal reconstruction chest CT of the same patient above, showing patchy ground-glass opacities.

Axial chest CT shows bilateral patchy consolidations (arrows), some with peripheral ground-glass opacity. Findings are in peripheral and subpleural distribution.

Axial chest CT of a COVID-19 patient demonstrating bilateral ground-glass opacities with round morphology.

Axial chest CT at lower level of the same patient above demonstrates more round morphology patchy ground-glass opacities.

COVID-19 axial chest CT shows bilateral patchy ground-glass opacities with septal thickening (crazy-paving).

Axial chest CT with right lower lobe opacities. Opacities with central consolidation and peripheral rim of ground-glass opacity (halo sign) can be seen in COVID-19.

Lesion in the peripheral right lower lobe of a COVID-19 patient demonstrates peripheral rim of consolidation with central ground-glass opacity (reversed halo sign).

Although chest x-ray has a lower sensitivity than chest CT in detection of COVID-19, findings on chest x-ray correlate with chest CT. Bilateral, lower lobe, and peripheral distribution of opacities are the most common expected findings on chest x-ray of COVID-19 cases.[5, 8]

(See the image below.)

Chest radiograph PA view demonstrates bilateral lower lobe and peripheral predominant hazy opacities with same distribution expected to be seen on chest CTs.

Although there have been few large studies on long-term lung radiographic changes in patients who have had COVID-19, there are studies that point to some possibly characteristic features, such as ground-glass opacities (GGOs) after 3 months, 6 months, or even a year.[10, 11, 12, 13, 14, 15, 16, 17]  In a study by Gao et al, GGOs were found to be highly prevalent at discharge (99.5%) in patients who were hospitalized for COVID-19, and GGOs were detected in 37.3% of patients at 3 months. The authors also noted that at 1 month after discharge, GGOs  and reticular pattern were most common in patients with mild and severe disease. At 3 months, parenchymal bands were present in 34.4%, thickening of adjacent pleura in 13.9%, and reticular lesions in 14.8%.[10]

In a prospective study, by Wu et al, of 83 patients with severe COVID-19 who were admitted to a hospital in Wuhan, China, but who did not require mechanical ventilation, 65 (78%) of patients had residual changes on CT at 3 months after discharge, with ground-glass opacity (65 [78%]), interlobular septal thickening (28 [34%]), reticular opacity (27 [33%]), and subpleural curvilinear opacity (9 [11%]) being the most common CT findings. At 6 months, GGO was still present in 38 (46%), interlobular septal thickening in 11 (13%), reticular opacity in 13 (16%), subpleural curvilinear opacity in 4 (5%), mosaic attenuation in 3 (4%), and bronchiectasis in 1 (1%).  At 9 months, the radiologic changes had not fully resolved in 22 patients (27%), primarily GGO (20 [24%]). There was no significant improvement at 12 months compared with 9 months, and at 12 months after discharge, the radiologic changes had not resolved fully in 24% of patients.[17]

In a study by Barini et al of 115 patients who had COVID-19 and received an initial chest CT scan in the ED and a follow-up chest CT scan 18 months after discharge, 6 patients with bilateral pneumonia at baseline had unilateral pneumonia at 18 months, and 19 patients had bilateral pneumonia both at baseline and at 18 months. In patients who showed bronchiectasis at baseline, 11 had complete resolution but 6 still had bronchiectasis at 18 months.[11]

Han et al, in a prospective study of  6-month follow-up chest CT scans of survivors of severe COVID-19 pneumonia, found that there was evidence of fibrotic-like changes in 40 of 114 patients (35%). According to the authors, the findings were associated with older age, acute respiratory distress syndrome, longer hospital stays, tachycardia, noninvasive mechanical ventilation, and higher initial chest CT score.[14]