Fig 1.
Flowchart for patient recruitment.
A database research encompassing the years 2008–2019 identified 1748 patients who underwent chest MRI. 128 patients had chest MRI in the setting of complicated pneumonia. Out of these, 12 patient were excluded due to incomplete lung MRI protocol or bad image quality. 33 patients had CXR within ±5 days of the baseline MRI scan and 12 patients had also CXR within ±5 days of the follow-up MRI scan.
Table 1.
Patient characteristics.
Table 2.
Intra-reader agreement for MRI and CXR scores.
Table 3.
Inter-reader agreement for MRI and CXR scores.
Fig 2.
Contrast-enhanced MRI, CXR and LUS study of a 3-year old child.
T2 weighted MRI (A) demonstrates an abscess formation within the consolidation in the middle lobe (red arrow), not detected by CXR (B) showing only consolidation (red arrow). In LUS (C) the abscess demarcates barley and was not diagnosed initially (red arrow).
Table 4.
Prevalence of morphologic findings in MRI and CXR at baseline.
Fig 3.
Contrast-enhanced T1 weighted sequence in comparison with T2 weighted sequence of a 7-year-old child.
Both MRI sequences clearly delineate an abscess formation in the left lower lobe. On T1-weighting imaging (A) the abscess is characterized by the missing central contrast uptake and peripheral rim enhancement (red arrow). On T2-weighting (b), the necrotic center of the abscess has a high signal intensity, whereas the thick capsule shows a lower intensity (red arrow). The corresponding CXR (c) shows a consolidation in the left lower lung field as well as the abscess (red arrow). The example shows that the extension of abscess formations might be underestimated in CXR.
Table 5.
Scores for morphologic findings in MRI and CXR at baseline.
Fig 4.
Comparison of T2-weighted sequences and CXR of a 9-year-old child at baseline and follow-up.
MRI shows an abscess formation in the right lower lobe at baseline (red arrow) (A), which has regressed under therapy at follow-up (red arrow) (B). The corresponding CXR (C) shows a slight reduction in transparency in the right lower lung field without any clear evidence of an abscess (red arrow). At follow-up, CXR (D) shows no pathological findings not delineating the full extent of residual inflammatory changes (red arrow).
Table 6.
Prevalence of findings in MRI and CXR at follow-up.
Fig 5.
Comparison of T2-weighted sequences and CXR of an 11-year-old child at baseline and follow-up.
MRI shows an abscess formation in the right lower lobe at baseline (red arrow) (A). The surrounding inflammatory changes regressed under therapy, whereas the abscess formation increased at follow-up (red arrow) (B). The abscess formation was not detected by CXR (red arrows) (C+D). Furthermore, the extension of pathologic findings was underestimated by CXR (D).
Table 7.
Differences in scores for morphologic findings in MRI and CXR between baseline and follow-up.
Fig 6.
Contrast-enhanced T1 weighted sequence in in a 7-year-old child with pleural effusion and a 6-year-old child with pleuritis.
Pleural effusion on the right side without pathologic contrast enhancement of the pleura (red arrow) (A). In comparison, also right sided pleural effusion, but with pathologic enhancement of the pleura, suggestive for an inflammatory process (red arrow) (B).