A Simplified CT Score for Thrombus Burden in Acute Pulmonary Embolism: Clinical Correlation and Reproducibility

Figures

• 

  Figure 1. Diagram of the GmScore for the simplified quantification of thrombus burden in acute pulmonary embolism. The score is based on the most proximal anatomical location of a thrombus causing ≥50% luminal occlusion: main pulmonary artery (GmS3), lobar arteries (GmS2), and segmental arteries (GmS1). The color coding (red, orange, and yellow, respectively) reflects the anatomical hierarchy and the estimated thrombus burden.

• 

  Figure 2. Flowchart of patient selection. Adult patients with acute pulmonary embolism confirmed by CT pulmonary angiography were consecutively identified from the institutional pulmonary embolism registry. CT examinations and clinical data were retrospectively reviewed. Recruitment for each anatomical category was closed once the predefined sample size of 30–40 patients per group had been reached, in accordance with the statistical requirements for association analysis and interobserver agreement estimation. All eligible cases were included in the final study cohort (n = 132). PE = Pulmonary embolism.

• 

  Figure 3. (A) Axial CT pulmonary angiography image showing right ventricular dilatation with an RV/LV ratio > 1 (4.56/1.49 cm = 3.06; black arrows), consistent with right ventricular strain. (B,C) Axial and sagittal images demonstrating a thrombus located in the right main pulmonary artery, occupying >50% of the vessel lumen (yellow circles). Findings correspond to GmS3. RV/LV = right ventricle to left ventricle diameter ratio.

• 

  Figure 4. Association between GmScore and clinical risk classifications. (a) Distribution of European Society of Cardiology (ESC) risk categories within each GmScore group. (b) Proportion of patients with simplified Pulmonary Embolism Severity Index (sPESI) ≥ 1 according to GmScore category.

• 

  Figure 5. Association between GmScore and the continuous RV/LV ratio. Points represent individual values (with slight horizontal jitter to avoid overlap), and boxplots display the median and interquartile range for each category (Spearman ρ = 0.75; 95% CI: 0.66–0.82; p < 0.001; n = 116). The solid line illustrates the overall trend. A progressive increase in the RV/LV ratio is observed with higher GmScore categories.