Titre
Diagnostic value of MRI for the detection of suspected placental invasion: Does it depend on observers' experience?
Type
poster de conférence/colloque
Institution
UNIL/CHUV/Unisanté + institutions partenaires
Auteur(s)
Schmidt, S.
Auteure/Auteur
Anaye, A.
Auteure/Auteur
Rey, A.
Auteure/Auteur
Alamo Maestre, L.
Auteure/Auteur
Denys, A.
Auteure/Auteur
Bongartz, G.
Auteure/Auteur
Terraz, S.
Auteure/Auteur
Artemisia, A.
Auteure/Auteur
Liens vers les personnes
Liens vers les unités
Titre du livre ou conférence/colloque
ECR 2011, European Annual Congress of Radiology
Adresse
Vienna, Austria, March 3-7, 2011
Statut éditorial
Publié
Date de publication
2011
Langue
anglais
Résumé
Purpose: To evaluate the diagnostic value of specific MR features for detection of suspected placental invasion according to observers' experience.Methods and Materials: Our study population included 25 pregnant women (mean age 35.16) investigated by prenatal MRI. In twelve out of them placental invasion was histopathologically proven, the 13 other women (52%) without placental invasion served as control group. Multiplanar T1- and T2-weighted sequences had been performed mostly without IV contrast injection (1.5 T). MR examinations of the two groups were rendered anonymous, mixed, then independently and retrospectively reviewed by two senior and two junior radiologists in view of 8 MR features indicating placentar invasion including the degree. Results were compared with surgical diagnosis (placenta normal/increta/accreta/percreta). Interobserver agrement between senior and junior readers were calculated. Stepwise logistic regression and receiver operating (ROC) curvers were performed.Results: Demographics between the two groups were not statistically different. Overall sensitivity and specificity for detecting placentar invasion was 90.9% and 75.0% for senior readers, and 81.8% and 61.8% for junior readers respectively. The most significant MR features indicating placentar invasion were T2 hypointense placental bands, followed by placenta praevia, focally interrupted myometrial border, posterior placental insertion, and heterogeneous placental signal. For each of the evaluated MR features the interobserver agreement kappa between the two senior readers was superior than that between the junior readers, ranging from bad (<0.4) to good (0.4-0.75).Conclusions: MRI can be a reliable and reproducible tool for detection of suspected placentar invasion, however very variable according to the observers' experience.
PID Serval
serval:BIB_90C86510FD23
Date de création
2012-03-13T17:09:43.241Z
Date de création dans IRIS
2025-05-20T22:02:44Z