Int J Clin Exp Med 2017;10(6):8917-8925www.ijcem.com /ISSN:1940-5901/IJCEM0051296

Original ArticleUltrasound and MRI accordance and features in the prenatal diagnosis of placenta accreta

Ru Lin1, Jingtao Li2, Yunyun Ren1, Haidong Cheng3

1Department of Ultrasound, Obstetrics and Gynecology Hospital of Fudan University, Shanghai 200011, China; 2Software School, Fudan University, Shanghai, China; 3Department of Obstetrics, Obstetrics and Gynecology Hos-pital of Fudan University, Shanghai 200011, China

Received January 18, 2017; Accepted June 5, 2017; Epub June 15, 2017; Published June 30, 2017

Abstract: Objectives: To analyze major image features of ultrasonography and magnetic resonance imaging (MRI), as well as the agreement between the two methods for the diagnosis of placenta accreta. Methods: A retrospective analysis was performed on 378 highly suspected cases of placenta accreta in Obstetrics and Gynaecology Hospi-tal of Fudan University, by comparing ultrasound and MRI results with postpartum surgical pathology and clinical follow-up results. Results: Clinical and postpartum pathology confirmed 65 cases of placenta accreta. Ultrasound and MRI results showed a high level of agreement (Kappa≈0.79). Ultrasound features with high accuracy (>60%) are as follows in the order of high to low: irregular blood flow signals in the basal portion of the placenta, intrapla-cental lacunae, placental thickening, thinning or disappearance of the hypoechoic retroplacental myometrial zone. The combination of the following three features has the largest Area Under Curve value (AUC, 0.93): Intraplacental lacunae, Thinning or disappearance of the hypoechoic retroplacental zone and Irregular blood flow signals in the basal portion of the placenta. MRI features with high accuracy (>60%) are as follows: Dark intraplacental bands on T2-weighted images, heterogeneous signal intensity within the placenta, blurring or disappearance of the low-signal-intensity interface between the placenta and the myometrium. The combination of all 6 features listed in this article has the largest AUC (0.79). Conclusion: Ultrasonography, with a high accuracy, is still the main method for the diagnosis of placenta accreta. Combining multiple image features can improve the accuracy of diagnosis for both ultrasound and MRI.

Keywords: Placenta accrete, ultrasound, magnetic resonance imaging, image features

Introduction

Placenta accreta refers to the reduction or absence of the deciduas, which might result from primary decidua dysplasia or traumatic endometrial inflammation; it characterized as the disappearance of fracture line of the stra-tum spongiosum between the placenta and the uterus, and one or more maternal cotyledons closely adhering to the basal layer of the decid-ua, or even invading or penetrating the myome-trium. Classified on the basis of the depth of myometrial invasion, there are placenta accre-ta vera, placenta increta and placenta percreta [1]. Studies have shown that factors that con-tribute to placenta accreta include uterine sur-geries like abortion and endometrial biopsy, advanced-age pregnancy and caesarean deliv-ery [1, 2].

Placenta accreta is a major cause of perinatal mortality. There has been a significant increase in the frequency of placenta accreta in recent years [3]. Accurate prenatal diagnosis of pla-centa accreta can help obstetricians to prepare in advance for hemorrhagic shock and the resuscitation of the newborn as well as to develop multiple effective treatment plans [4, 5].

Ultrasound and MRI are important diagnostic methods of placenta accreta. However, there is no consistent conclusion about their image fea-tures and the accuracy of diagnosis corre-sponding to each feature, and it’s not clear whether there is an agreement between the diagnoses obtained using the two methods. In this study, we used the comprehensive clinical and surgical pathology diagnosis as the gold

Ultrasound and MRI features in placenta accreta

8918 Int J Clin Exp Med 2017;10(6):8917-8925

Table 1. Basic information and clinic featuren=378

Average age (in years) 31.9±4.3Gestational age at delivery (in weeks) 36.5±3.63Gravidity 3.33±1.85Parity 1.64±1.46Medical history Prior medical abortion/Uterine curettage 261 (69%) Prior caesarean section 102 (30%)Average gestational age at the time of diagnosis by ultrasonography (in weeks) 23Average gestational age at the time of MRI (in weeks) 33Placenta implantation site Anterior wall 147 (39%) Posterior wall 109 (29%) Lateral wall/lower part of uterus 122 (32%)Delivery mode Vaginal delivery 173 (46%) Elective caesarean section 205 (54%)Amount of bleeding <500 ml 239 (63%) ≥500 ml and <2000 ml 112 (30) ≥2000 ml 27 (7%)Final diagnosis Placenta accreta vera 17 Placenta increta/percreta 48 Non-adherent placenta 313

standard to examine the agreement of prenatal ultrasound and MRI diagnosis of placenta accreta, and to identify the most relevant abdominal ultrasound image features and MRI scanned image features for confirming placen-ta accreta.

Materials and methods

Subjects

A retrospective analysis was performed on 378 high-risk suspected cases of placenta accreta with placenta previa, prior caesarean delivery, prior uterine surgery and prior abortion at the Obstetrics and Gynaecology Hospital of Fudan University from May 2012 to September 2016. All cases received ultrasound and low-field MRI examination before delivery and were followed up till natural or caesarean delivery. Medical chart review was used to record clinical infor-mation (Table 1). The suspected depth of myo-metrial involvement did not affect counselling or clinical management [6], and for the purpos-

es of this study, the diagnosis of placenta accreta refers to placenta increta and percreta as well as accreta vera.

Instruments and examination procedures

Ultrasound examination: The equipment used were GE Voluson E8 (GE Medical Systems, Kretz Ultrasound, Zipf, Austria) and Philips HD11XE color Doppler ultrasound system (Philips Healthcare, Eindhoven, Netherland) with a convex array probe of 3.5 MHz frequen-cy. Pregnant women were asked to properly fill their bladder before receiving the abdominal ultrasound examination. After routine examina-tion of the fetus, radiologists used a 2-dimen-sional color Doppler to observe the appearance of the lower uterine segment and the continuity of the myometrium; Also observed were the location of the placenta, placental thickness, internal echo of the placenta, appearance of the vessels at placenta implantation site, appearance and echo of the myometrium behind the implantation site as well as the

Ultrasound and MRI features in placenta accreta

8919 Int J Clin Exp Med 2017;10(6):8917-8925

Figure 3. Thickening of the arcuate arteries behind the placenta. The vessels are shown to be tortu-ous with abundant blood flow, making an irregular-shaped mass (arrow). P, placenta; B, bladder.

Figure 1. Unclear boundaries between the placenta and the uterus (two arrows), irregular intraplacental lacunae (arrow). P, placenta; B, bladder.

blood flow in the arcuate arteries. Suspicious placenta implantation areas were identified.

MRI examination: Pregnant women entered the equipment headfirst, lying on their back or side and breathing calmly. T2-weighted images were acquired in the coronal, sagittal and axial planes of the middle and lower abdomen and the pelvis; T1-weighted images were acquired in the axial plane; Diffusion weighted images were acquired in the axial plane. Scan series were acquired using a Siemens Avanto 1.5T superconductive MRI scanner (Siemens Healthcare GmbH, Erlangen, Germany) with a 6-channel phased-array coil.

Diagnostic criteria

Placenta accreta was defined by clinical criteria at the time of delivery and by pathologic find-ings. Ultrasound and MRI examinations were conducted and their results reviewed by 2 sonographers and 2 radiologists respectively, both of whom had over 5 years of experience working on diagnosing placenta abnormalities. They were blinded to the clinical diagnosis and were asked to mark image features based on the diagnostic criteria below and provide inter-pretations. The interpretations can be classi-fied into 3 types: 1) conclusive; 2) suggestive; 3) negative. Two sonographers or radiologists needed to provide one final interpretation agreed by both after discussion when there was a disagreement. The result was considered to be positive if the final interpretation provided by the two sonographers was suggestive of, or conclusive for placenta accreta. The MRI result was considered to be positive if the final inter-pretation provided by the two radiologists was suggestive or conclusive.

Ultrasound diagnostic criteria: The ultrasound features considered to be consistent with pla-centa accreta included the following: (1) Placental thickening; (2) Intraplacental lacu-nae: visualization of multiple liquid dark areas (vein lakes) in various sizes and shapes in the placenta (Figure 1); (3) Thinning or disappear-ance of the retroplacental hypoechoic zone: when the sound beam is aligned with the tan-gent line, the myometrial hypoechoic zone at

Figure 2. Irregular intraplacental lacunae (arrow); disappearance of the retroplacental hypoechoic zone, thinning of the myometrium (cross). Chorionic villi invasion of the uterus serosa confirmed at deliv-ery. P, placenta.

Ultrasound and MRI features in placenta accreta

8920 Int J Clin Exp Med 2017;10(6):8917-8925

Table 3. Agreement of Ultrasound (US) and MRI Diagnostic Results

MRIKappa

+ –US + 59 2 0.79

– 22 295

Table 2. Agreement of Ultrasound (US) and MRI Results with Comprehensive Clinical and Pathological Results

Gold standardKappa

+ –US + 55 6 0.85

– 10 307MRI + 58 23 0.75

– 7 290

Figure 4. Both the MRI sagittal image (A) and coronal image (B) show local thinning of the myometrium and local invasion of placental tissues into the myometrium (arrows). Placental implantation into the lower myometrium of the anterior uterine wall confirmed at delivery. P, placenta.

placental attachment site is shown as thinned or absent (Figure 2); (4) Discontinuity in the bladder serosa: when the implanted placenta penetrates the myometrium and reach the serosa, if the implantation site is at the anterior uterine wall and behind the bladder, the hyper-echoic band of the bladder serosa adjacent to the uterus disappears and irregular anechoic structures protruding into the bladder can also be seen; (5) Irregular blood flow signals in the

ages show intraplacental low-signal-intensity bands; (4) Heterogeneous signal intensity in the placenta; (5) Multiple thickened tortuous vessels in the placenta; (6) Bulging of the lower uterine segment.

Comprehensive clinical and pathological crite-ria: In placenta accreta, the placenta is closely adherent to the endometrium and surgically uncontrollable hemorrhage occurs when attempted to remove; In placenta increta, the placental tissues that have invaded the myo-metrium have to be removed through curet-tage; In placenta percreta, it is visible to the naked eye that the placental tissues have invaded the entire endometrium and even the adjacent organs outside of the uterus [8].

Statistical analysis

The data in this study were processed and ana-lyzed with the Stata12.0 (StataCorp, College Station, TX, USA). Compared against the gold standard of comprehensive clinical and patho-logical diagnosis, we evaluated the sensitivity, specificity, positive predictive value, negative predictive value and accuracy of ultrasound and MRI features for the diagnosis of placenta accreta. The Kappa statistic was used to evalu-ate the agreement of MRI and ultrasound

basal portion of the placen-ta: the blood flow of the arcuate arteries in the ret-roplacental myometrium is shown as interrupted, ab- sent or an irregular-shaped mass (Figure 3).

MRI diagnostic criteria: Currently, there are no con-sistent MRI diagnostic cri-teria for placenta accreta. Lim PS [7] has reported the following diagnostic fea-tures (Figure 4A, 4B): (1) Local blurring and disap-pearance of the low-signal-intensity band between the placenta and the myome-trium; (2) Local protrusion of the high-signal-intensity placenta tissues into the low-signal-intensity myome- trium; (3) T2 weighted im-

Ultrasound and MRI features in placenta accreta

8921 Int J Clin Exp Med 2017;10(6):8917-8925

results; The chi-square test was used to evalu-ate the difference of the MRI and ultrasound results as well as each ultrasound and MRI feature. P<0.05 was considered statistically significant.

Results

Sixty-five women were ultimately found to have placenta accreta and were confirmed through clinical or surgical pathology. Clinical informa-tion is shown in Table 1.

Agreement between US and MRI

The diagnosis was correct in 95.8% of cases with ultrasonography and in 92.1% with MRI. Agreement of ultrasound and MRI results with comprehensive clinical and pathological results is shown in Table 2. There was a high level of agreement between ultrasound and clinical results (Kappa≈0.85). There was also a high level of agreement between MRI and clinical results (Kappa≈0.75). Agreement of ultrasound and MRI results for the diagnosis of placenta accreta is shown in Table 3. There was a high level of agreement between the two diagnostic methods (Kappa≈0.79). There was a statisti-cally significant difference between the posi-tive rate of ultrasound and MRI tests (x2≈15, P <0.05).

Ultrasound and MRI features

In order to identify the ultrasound and MRI fea-tures with the highest accuracy, we re-evaluat-ed the ultrasound and MRI images with at least one indicator of placenta accreta (n=83). The sensitivity, specificity and accuracy of ultra-sound and MRI features are summarized in Table 4.

Four image features of placenta accreta were significantly different from those of a normal placenta (P<0.01). They are as follows: Discontinuity in the bladder serosa; Local pro-trusion of high-signal-intensity placental issues into the low-signal-intensity myometrium; Multiple thickened tortuous vessels in the pla-centa; Uterine bulging.

Ultrasound image features identified are as fol-lows in the order of high to low accuracy: Irregular blood flow signals in the basal portion of the placenta (91.6%), Intraplacental lacunae (88%), placental thickening (83.1%), thinning or disappearance of the hypoechoic retroplacen-tal myometrial zone (75.9%). Irregular blood flow signals in the basal portion of the placenta had the highest sensitivity (91.4%) with a speci-ficity of 92%; Discontinuity in the bladder sero-sa had the best specificity (100%), but a low sensitivity (36.2%).

Table 4. Ultrasound and MRI features

n=83 TP FP P Sen Spe PPV NPV AcUltrasound Features Intraplacental lacunae 52 4 0.752 89.7 84.0 92.9 77.8 88.0 Thinning or disappearance of the hypoechoic retropla-cental zone

44 6 0.118 75.9 76.0 88.0 57.6 75.9

Placental thickening 46 3 0.061 81.0 88.0 94.0 66.7 83.1 Irregular bloodflow signals in the basal portion of the placenta

53 2 0.450 91.4 92.0 96.4 82.1 91.6

Discontinuity in the bladder serosa 21 0 <0.01 36.2 100.0 100.0 40.3 55.4MRI Features Local blurring or disappearance of the low-signal-intensi-ty interface between the placenta and the myometrium

48 20 0.100 82.8 20.0 70.6 33.3 63.9

Local protrusion of the high-signal-intensity placental tissues into the low-signal-intensity myometrium

17 1 <0.01 29.3 96.0 94.4 36.9 49.4

Dark intraplacental bands on T2-weighted images 45 12 1.000 77.6 52.0 78.9 50.0 69.9 Heterogeneous intraplacental signal intensity 39 8 0.054 67.2 68.0 83.0 47.2 67.5 Multiple thickened tortuous vessels in the placenta 12 4 <0.01 20.7 84.0 75.0 31.3 39.8 Uterine bulging 19 4 <0.01 32.8 84.0 82.6 35.0 48.2TP=True positives, FP=False positives, Sen=sensitivity, Spe=specificity, PPV=positive predictive value, NPV=negative predictive value, Ac=Accuracy.

Ultrasound and MRI features in placenta accreta

8922 Int J Clin Exp Med 2017;10(6):8917-8925

Figure 5. ROC for each ultrasound feature and combinations of multiple fea-tures. IL, Intraplacental lacunae; TDHR, Thinning or disappearance of the hy-poechoic retroplacental zone; PT, Placental thickening; IBS, Irregular blood flow signals in the basal portion of the placenta; DBS, Discontinuity in the bladder serosa; Traditional, Diagnosis made based on the experience of so-nographers.

MRI features identified are as follows in the order of high to low accuracy: Dark intraplacen-tal bands on T2-weighted images (69.9%), het-

The ROC results of each MRI feature and their combinations are shown in Figure 6. The upper left point (with the highest Youden index and

Figure 6. ROC for each MRI image feature and combinations of multiple fea-tures. LBD, Local blurring or disappearance of the low-signal-intensity inter-face between the placenta and the myometrium; LP, Local protrusion of the high-signal-intensity placental tissues into the low-signal-intensity myome-trium; DIB, Dark intraplacental bands on T2-weighted images; HI, Heteroge-neous intraplacental signal intensity; MTTV, Multiple thickened tortuous ves-sels in the placenta; UB, Uterine bulging; Traditional, Diagnosis made based on the experience of radiologists.

erogeneous signal intensity within the placenta (67.5%), local blurring or disappear-ance of the low-signal-inten-sity interface between the placenta and the myometri-um (63.9%). Local blurring or disappearance of the low- signal-intensity interface be- tween the placenta and the myometrium had the best sensitivity (82.8%) but a low specificity (20%). Local pro-trusion of high-signal-intensi-ty placental issues into the low-signal-intensity myome-trium had the best specificity (96%) but a low sensitivity (29.3%).

Using the comprehensive clinical and surgical patholo-gy diagnosis as the gold stan-dard, we performed Recei- ver Operating Characteristics (ROC) analysis on each ultra-sound image feature and vari-ous combinations of multiple features. The upper left point (with the highest Youden index and maximum distance from the reference line) was selected as the cut-off value (Figure 5). The combination of the following three features has the largest Area Under Curve value (AUC, 0.93): Intraplacental lacunae, Thin- ning or disappearance of the hypoechoic retroplacen-tal zone and Irregular blood flow signals in the basal por-tion of the placenta. This value is higher than the AUC of any single feature (maxi-mum 0.916) as well as the AUC (0.8541) of the tradition-al diagnosis method based on the experience of sono- graphers.

Ultrasound and MRI features in placenta accreta

8923 Int J Clin Exp Med 2017;10(6):8917-8925

maximum distance from the reference line) was selected as the cut-off value. The combination of all 6 features listed in this article has the largest AUC (0.7917). This result is better than the AUC (maximum 0.6762) of any single fea-ture as well as the AUC (0.54) of the traditional diagnosis method based on the experience of radiologists.

Discussion

Because ultrasound is relatively economical and easy to implement, it is still the preferred method for the diagnosis of placenta accreta. There is still controversy over the safety of MRI examination for pregnant women and fetuses. It is mainly used when the implantation site is beyond the field of view of ultrasound and diffi-cult to image, or as an auxiliary and comple-mentary technique when ultrasound results suggest possible placenta accreta [9]. The American College of Radiology and the European Medicines Agency do not recom-mend having enhanced MRI scans during preg-nancy unless necessary [10, 20]. The Food and Drug Administration of United State does not recommend having MRI scans in the first 3 months of pregnancy. Some studies suggested that except for the first 3 months, patients can have MRI scans any time during their pregnan-cy and that the technique is relatively safe for pregnant women and fetuses [11, 12]. And with its multidimensional imaging, good resolution for soft tissues and better ability to identify abnormal placenta and its depth of implanta-tion without the influence of patient’s weight and the position of the placenta, MRI can be used as a regular examination method [13].

Previous studies have reported different results about the sensitivity and specificity of ultra-sound and MRI for the diagnosis of placenta accreta. Some studies suggested that the dif-ference between the sensitivity and specificity of these two diagnostic methods was statisti-cally significant [15]. Other studies concluded that the difference was not statistically signifi-cant [16]. A systematic review showed that the sensitivity of ultrasound was 77%-88% and the specificity was 93%-96%, the sensitivity of MRI was 72%-90% and the specificity was 81%-94%, and that the difference was not statisti-cally significant [17]. In this study, the sensitivi-

ty and specificity of ultrasound for the diagnosis of placenta accreta were 84.6% and 98.1% respectively, and those of MRI were 89.2% and 92.7% respectively. The results were consistent with what the systematic review reported [17]. Both the ultrasound sensitivity and specificity for placenta accreta were higher than the ultra-sound sensitivity (74%-81%) and specificity (93%-95%) reported by Lu M [19]. In this study, we found a good agreement between the diag-nostic results of the two methods using the Kappa statistic. There was a statistically signifi-cant difference between the positive rate for placenta accreta of the two methods.

In this study, the false negative rate of ultra-sound for placenta accreta (15.4%) was slightly higher than that of MRI (10.8%). An analysis of cases false-negatively diagnosed by ultrasound but detected by MRI showed that in these cases, the placenta implanted on the posterior uterine wall and the ultrasound was not able to identify the depth of implantation [18]. Also the false positive rate of MRI examination for pla-centa accreta in this study was slightly high (7.3%). The positive predictive value was 71.6%. As in the literature [8, 14, 17, 21], we found that the best PPV (90%) of MRI occurred when intra-placental T2 dark bands were associated with the disappearance of the myometrium and uterine bulging. After a retrospective analysis of the 23 MRI false-positively-diagnosed cases, we found that 20 of the cases showed blurring of the placenta-myometrium interface, local thinning of the myometrium, intraplacental T2 dark bands and tortuous blood vessels. A diag-nosis of placenta invasion was made for these cases. The misdiagnosis may have been caused by placenta enlargement and myome-trium thinning during middle and late pregnan-cy, making it difficult to observe the dark band of the placenta-myometrium interface. A retro-spective analysis of the 23 misdiagnosed MRI cases showed that 18 cases had homogeneous intraplacental signal intensity, 17 cases had few T2 dark bands and tortuous vessels, 13 cases showed no local thickening of tortuous vessels between the placenta and myometri-um. Therefore, radiologists should carefully examine the interface of the placenta and myo-metrium in patients during middle and late pregnancy and pay attention to indirect signs in order to reduce misdiagnosis.

Ultrasound and MRI features in placenta accreta

8924 Int J Clin Exp Med 2017;10(6):8917-8925

In summary, there is a high level of agreement between ultrasound and MRI in ruling out the diagnosis of placenta accreta in patients at risk. Ultrasound is relatively economical and easy to implement; it is still the preferred method. MRI, as a complementary technique, appears to be helpful when the patients consid-ered to be suggestive but not conclusive at ultrasound, especially when the placenta implants on the posterior uterine wall. In these situations, MRI can be used to confirm or refute the diagnosis of placenta accreta. Combining multiple image features can improve the accu-racy of diagnosis for both ultrasound and MRI.

Disclosure of conflict of interest

None.

Address correspondence to: Haidong Cheng, De- partment of Obstetrics, Obstetrics and Gynecology Hospital of Fudan University, 419 Fangxie Road, Shanghai 200011, China. Tel: (86) 21-33189900, E-mail: hdcheng1@yeah.net

References

[1] Wortman AC, Alexander JM. Placenta accreta, increta, and percreta. Obstet Gynecol Clin North Am 2013; 40: 137-154.

[2] Wu S, Kocherginsky M, Hibbard JU. Abnormal placentation: twenty-year analysis. Am J Obstet Gynecol 2005; 192: 1458-1461.

[3] Seet EL, Kay HH, Wu S, Terplan M. Placenta accreta: depth of invasion and neonatal out-comes. J Matern Fetal Neonatal Med 2012; 25: 2042-2045.

[4] Robinson BK, Grobman WA. Effectiveness of timing strategies for delivery of individuals with placenta previa and accreta. Obstet Gynecol 2010; 116: 835-842.

[5] Tikkanen M, Paavonen J, Loukovaara M, Stefa-novic V. Antenatal diagnosis of placenta ac-creta leads to reduced blood loss. Acta Obstet Gynecol Scand 2011; 90: 1140-1146.

[6] Warshak CR, Eskander R, Hull AD, Scioscia AL, Mattrey RF, Benirschke K, Resnik R. Accuracy of ultrasonography and magnetic resonance imaging in the diagnosis of placenta accreta. Obstet Gynecol 2006; 108:573-581.

[7] Lim PS, Greenberg M, Edelson MI, Bell KA, Ed-monds PR, Mackey AM. Utility of ultrasound and MRI in prenatal diagnosis of placenta ac-creta: a pilot study. AJR Am J Roentgenol 2011; 197: 1506-1513.

[8] Derman AY, Nikac V, Haberman S, Zelenko N, Opsha O, Flyer M. MRI of placenta accreta: a

new imaging perspective. Am J Roentgenol 2011; 197: 1514-1521.

[9] Mar WA, Berggruen S, Atueyi U, Sekhon S, Gar-zon SA, Knuttinen MG, McGahan JP. Ultra-sound imaging of placenta accreta with MR correlation. Ultrasound Quarterly 2015; 31: 23-33.

[10] Committee on Obstetric Practice. Committee opinion no. 529: placenta accreta. Obstet Gy-necol 2012; 120: 207-211.

[11] Teo TH, Law YM, Tay KH, Tan BS, Cheah FK. Use of magnetic resonance imaging in evalua-tion of placental invasion. Clin Radiol 2009; 64: 511-516.

[12] Alamo L, Anaye A, Rey J, Denys A, Bongartz G, Terraz S, Artemisia S, Meuli R, Schmidt S. De-tection of suspected placental invasion by MRI: do the results depend on observer’ expe-rience. Eur J Radiol 2013; 82: 51-57.

[13] Palacios Jaraquemada JM, Bruno CH. Magnet-ic resonance imaging in 300 cases of placenta accreta: surgical correlation of new findings. Acta Obstet Gynecol Scand 2005; 84: 716-724.

[14] Lax A, Prince MR, Mennitt KW, Schwebach JR, Budorick NE. The value of specific MRI fea-tures in the evaluation of suspected placental invasion. Magn Reson Imaging 2007; 25: 87-93.

[15] Maher MA, Abdelaziz A, Bazeed MF. Diagnostic accuracy of ultrasound and MRI in the prena-tal diagnosis of placenta accreta. Acta Obstet Gynecol Scand 2013; 92: 1017-1022.

[16] Peker N, Turan V, Ergenoglu M, Yeniel O, Sever A, Kazandi M, Zekioglu O. Assessment of total placenta previa by magnetic resonance imag-ing and ultrasonography to detect placenta ac-creta and its variants. Ginekol Pol 2013; 84: 186-192.

[17] Meng X, Xie L, Song W. Comparing the diagnos-tic value of ultrasound and magnetic reso-nance imaging for placenta accreta: a system-atic review and meta-analysis. Ultrasound Med Biol 2013; 39: 1958-1965.

[18] Masselli G, Brunelli R, Casciani E, Polettini E, Piccioni MG, Anceschi M, Gualdi G. Magnetic resonance imaging in the evaluation of placen-tal adhesive disorders: correlation with color Doppler ultrasound. Eur Radiol 2008; 18: 1292-1299.

[19] Lu M, Wang Z, Chen C, Sun B. Accuracy of ul-trasound in the diagnosis of placenta accreta: an updated meta-analysis. Int J Clin Exp Med 2016; 9: 8887-8897.

[20] Expert Panel on MR Safety, Kanal E, Barkovich AJ, Bell C, Borgstede JP, Bradley WG Jr, Froelich JW, Gimbel JR, Gosbee JW, Kuhni-Kaminski E, Larson PA, Lester JW Jr, Nyenhuis J, Schaefer

Ultrasound and MRI features in placenta accreta

8925 Int J Clin Exp Med 2017;10(6):8917-8925

DJ, Sebek EA, Weinreb J, Wilkoff BL, Woods TO, Lucey L, Hernandez D. ACR guidance docu-ment on MR safe practices: 2013. J Magn Re-son Imaging 2013; 37: 501-530.

[21] D’Antonio F, Iacovella C, Palacios-Jaraquema-da J, Bruno CH, Manzoli L, Bhide A. Prenatal

identification of invasive placentation using magnetic resonance imaging: systematic re-view and meta-analysis. Ultrasound Obstet Gy-necol 2014; 44: 8-16.