Purpose or Case Report: Congenital Pulmonary Airway Malformation Volume Ratio (CVR) is widely used during prenatal ultrasound to predict outcome of fetuses with bronchopulmonary malformations (BPM). Since fetal MRI is often performed in fetuses with BPM, the aim of our study is to correlate CVR measurements by MRI and US, to compare inter-observer agreement of CVR measured by area-based MRI volume, diameter-based volume and to assess whether large lesions would be prone to more variability in CVRs among observers.
Methods & Materials: Under IRB approval, a retrospective analysis (2007-2016) of 51 patients who underwent US and fetal MRI within 24 hours was performed by 3 pediatric radiologists experienced in fetal imaging. Lesion volumes by diameter (US and MRI), lesion volume by area (MRI) were measured. A free-hand contouring tool was used on MRI to measure the lesion volume by area, which was then summed and divided by slice-thickness. Lesion volume by diameter both on US and MRI was calculated as [lesion length x height x width x 0.52]. The CVR for MRI and US was based on the lesion volume/head circumference with head circumference obtained from the same day US. The choice of the planar reformat for the measurement was left to the reviewers’ discretion. One radiologist measured the US CVR. The 3 MRI observers were blinded from the other observers’ measurements. Intraclass correlation (ICC) was performed.
Results: There was excellent inter-observer agreement of the CVR MRI measurements using both diameter and area methods with respectively ICC of 0.97 (95% CI=0.96-0.98) and ICC of 0.96 (95% CI=0.93-0.97). With larger lesion size, there was a greater variation of CVR and lower inter-observer agreement for both MRI methods. Inter-observer variation in larger lesions was independent of the method used. CVR by diameter on MRI to US CVR showed an ICC range of 0.84-0.92 and CVR by area on MRI to US CVR showed an ICC range of 0.84-0.94. In 2/3 observers, CVR by diameter on MRI better correlated with US CVR. There was no consistent trend that CVR on US was reliably larger or smaller than CVR obtained by either of the MRI CVR methods.
Conclusions: Our study shows that CVR measured on MRI using either diameter or area has excellent inter-observer agreement and correlate equally with CVR by US. Our data promotes CVR measurements based on diameter rather than more time-consuming area measurements. Larger lesions have less reproducibility in the CVRs both on US and MRI methods.