Pattern recognition of high-resolution computer tomography (HRCT) chest to guide clinical management in patients with mild to moderate COVID-19

Abstract

Aim: To describe the distribution of lung patterns determined by High Resolution Computed Tomography (HRCT) in COVID patients with mild and moderate lung involvement and outcomes after early identification and management with steroids and anticoagulants.

Material and methods: A cross sectional study of COVID-19 patients with mild and moderate lung involvement presenting at 5 healthcare centres in Trichy district of South TamilNadu in India. Patients underwent HRCT to assess patterns and severity of lung involvement, Inflammatory markers (LDH/Ferritin) and D-Dimer assay and clinical correlation with signs and symptoms. Patients were assessed for oxygen, steroid and anticoagulant therapy, clinical recovery or progression on follow up and details on mortality were collected. The RSNA, Fleischer Society guidelines and CORADS score was used for radiological reporting. New potential classification of patterns of percentage of lung parenchyma involvement in Covid patients is being suggested.

Results: The study included 7,340 patients with suspected COVID and 3,963 (53.9%) patients had lung involvement based on HRCT. RT PCR was positive in 74.1% of the CT Positive cases. Crazy Pavement pattern was predominant (n = 2022, 51.0%) and Ground Glass Opacity (GGO) was found in 1,941 (49.0%) patients in the study. Severe lung involvement was more common in the Crazy Pavement pattern. Patients with GGO in moderate lung involvement were significantly more likely to recover faster compared to Crazy Pavement pattern (P value <0.001).

Conclusion: HRCT chest and assessment of lung patterns can help triage patients to home quarantine and hospital admission. Early initiation of steroids and anticoagulants based on lung patterns can prevent progression to more severe stages and aid early recovery. HRCT can play a major role to triage and guide management especially as RT PCR testing and results are delayed for the benefit of patients and in a social cause to decrease the spread of the virus.

Keywords: COVID 19; HRCT; crazy pavement pattern; ground glass opacity.

Figure 1 (A and B)

A and B: 3D segmentation Volume images showing Mild and Moderate lung involvement, for segmentation and percentage analysis with the software

Figure 2

Volume of lung involvement by segmentation, Qualification and Analysis

Figure 3 (A and B)

A and B: Severe Disease by coronal CT & Segmentation analysis A 62 year-old male 75% lung involvement - “M” pattern, hospitalised with breathlessness SaO2 85% in room air with elevated inflammatory markers and D-Dimer

Figure 4 (A and B)

A and B: Axial and Coronal - Crazy paving pattern - moderate disease (CORADS -5 - Moderate Disease -”C” Pattern

Figure 5 (A and B)

A and B Anatomy of the Lobule of Lung and Pattern of Lung involvement : Secondary lobule is made up of Centrilobular structures, Lobular parenchyma and the septal structures. Secondary pulmonary lobules represent a cluster of up to 30 acini[9] (Light blue) supplied by a common distal pulmonary artery (blue) and bronchiole.[101112] These clustered acini are bounded by interstitial fibrous septa (interlobular septa), which has lymphatics (yellow) and Pulmonary veins (red) which outline an irregular polyhedron of varying size between 1 and 2.5 cm. Peripheral lobules are larger and cuboidal/pyramidal, while central lobules tend to be smaller and hexagonal[12] When the airspace is involved the HRCT reveals ground glass opacities (GGO), with dilated vessel sign and when the inflammation progresses through the lobular parenchyma and involves the septum, crazy-paving pattern is seen, which is due to interstitial vasculitis involving the pulmonary vein (box with arrow)

Figure 6

Ground glass opacities in posterior segments of bilateral upper lobes with dilated terminal pulmonary artery (blue double head arrow) - (due to airspace inflammation - Lung response to the Viral disease)

Figure 7

Crazy Paving and areas of consolidation

Figure 8

Crazy paving pattern

Chart 1

Pattens of Lung involvement, its pathophysiology and correlation with Biochemical parameters

Chart 2

Patterns in cases with Less than 30% Lung involvement on HRCT by Covid-19 with investigation and management protocol

Chart 3

Patterns in cases with Less than 30-60% Lung involvement on HRCT by Covid-19 with investigation and management protocol