Comparison of T2-weighted and diffusion-weighted imaging for the diagnosis of placenta accreta spectrum abnormality

Abstract

Background

Diffusion-weighted imaging (DWI) is feasible in prenatal imaging, and it exhibits better contrast between the placenta and the myometrium compared to T2-weighted (T2W) images.

Purpose

To compare magnetic resonance imaging (MRI) features of placenta accreta on T2W and DW imaging.

Material and Methods

In this retrospective study, 42 pregnant patients who underwent prenatal MRI were included. MRI was performed on a Siemens 1.5-T scanner. T2W and DWI sequences in the axial, sagittal, and/or coronal planes were compiled for review. Two radiologists independently interpreted T2W and DW images for placenta accreta. T2W and DWI scores were calculated based on the presence of features and graded as low, intermediate, and high risk. The association between imaging features and placental invasion on pathology was calculated using chi-square tests. Sensitivity, specificity, and positive and negative predictive values (NPV) were compared between T2W and DWI interpretations. Inter-reader agreement between the two radiologists for T2W and DWI scores was calculated using Cohen's kappa coefficient.

Results

Out of 42 pregnant patients, 10 were pathologically/surgically proven to have placenta accreta. There were no significant differences between T2W and DWI interpretations. Considering a cutoff >6 as positive, the T2W score had higher sensitivity (90% vs. 80%) and NPV (96.9% vs. 94.1%) than the DWI score. The specificity and positive predictive value were 100% for both scores. The inter-reader agreement of T2W score was higher (k = 0.943 vs. 0.882).

Conclusion

T2W and DWI are comparable in diagnosing placenta accreta spectrum. T2W sequences have higher sensitivity, NPV, and inter-reader agreement than DWI.

Keywords

Placenta accreta placental invasion prenatal magnetic resonance imaging diffusion-weighted imaging

Introduction

Placenta accreta spectrum abnormality refers to a condition in which the placenta fails to separate partially or completely from the decidua basalis during labor. This occurs due to abnormal attachment of the trophoblastic villi to the myometrium (1). Placenta accreta spectrum ranges from placenta accreta, in which the placenta is adhered to the myometrium, to placenta increta, in which the placenta grows into the myometrium, to placenta percreta, in which the placenta grows up to the uterine serosa. It may also extend into the extrauterine tissues such as the bladder, bowel, and mesentery in its most severe form (2). In this article, the term placenta accreta refers to this entire spectrum.

Prenatal diagnosis of placenta accreta is essential as failure to recognize this condition, and any attempt to forcefully separate the placenta during labor, may result in life-threatening hemorrhage and even maternal and fetal death (3). The risk of maternal and fetal mortality has been reported to be around 6%–7% and 9%–19%, respectively, in the setting of placenta accreta (4 –6). Placenta previa and history of previous cesarean delivery are the two most important risk factors for placenta accreta. A history of previous cesarean delivery increases the risk of placenta accreta by 8.7-fold (7). A combination of placenta previa and a history of three or more cesarean deliveries increases the risk to 67% (8). Due to the increasing number of cesarean deliveries, the incidence of placenta accreta has also increased in the past few decades. Other risk factors include a history of dilatation and curettage and previous uterine surgery such as myomectomy. The management of placenta accreta requires a multidisciplinary approach, including a team of obstetricians, radiologists, and surgeons, as well as counseling for future fertility options (9).

Ultrasound is the first line of investigation for the diagnosis of placenta accreta. A meta-analysis of prenatal ultrasound diagnosis of placenta accreta, including 14 cohort studies of 3209 pregnancies, found it to be highly sensitive and specific (10). However, the results vary between studies. This is due to the high operator dependence of ultrasound (11).

Magnetic resonance imaging (MRI) is often requested in suspected cases of placenta accreta, especially when the ultrasound findings are equivocal, in cases of posterior placenta, and to determine the extent of extrauterine tissue involvement in cases of placenta percreta. T2-weighted (T2W) images have been used conventionally for assessing placental invasion. However, recently diffusion-weighted imaging (DWI) has been shown to be better in delineating the placental myometrial interface due to the high contrast between the placenta and myometrium on DWI (12).

Many signs of placenta accreta have been described on MRI (13). However, no sign is pathognomonic, and sometimes these signs may also be seen in normal cases. Therefore, emphasis should be placed on cumulative scores based on clinical risk factors, ultrasound findings, and the presence of multiple MRI signs instead of relying on a single sign for the diagnosis of placenta accreta.

To our knowledge, there has been no previous study comparing T2W and DWI sequences for the diagnosis of placenta accreta. In addition, the application and comparison of a comprehensive scoring system on both T2W and DWI sequences has not been evaluated previously.

Material and Methods

Study cohort

The present study was approved by our institutional review board and by the ethical review committee of our institute. This was a retrospective study. Our institute has been regularly using DWI in prenatal MRI. All singleton pregnant patients who were referred for MRI for various indications from August 2021 to July 2022 (n = 55) were reviewed. The various indications for referral to MRI included suspicion of placenta accreta on ultrasound (n = 49), acute abdominal pain (n = 4), and fetal anomalies (n = 2). After applying the exclusion criteria, 42 patients were included in the study. The reasons for exclusion included lack of DWI sequence in the scan (n = 3), lack of pathological data (n = 8), lack of proper ultrasound report (n = 1), and poor quality of images (n = 1) (Fig. 1). All 42 patients included in our study had delivered at our own institute. The clinical details and ultrasound report were separately documented for each patient. Anonymized T2W and DWI with ADC images of all the included patients were compiled into two separate sets of digital versatile discs (DVDs).

MRI protocol

All MR examinations were performed on a Siemens 1.5-T Magnetom Avanto (Siemens Healthcare, Erlangen, Germany). The patient was placed in the supine position, and a phased array pelvic coil was used. Efforts were made to scan while the urinary bladder was mildly distended. Overdistension was avoided. The T2W and DWI sequences were obtained using uniform parameters for all the patients.

T2W images were acquired using HASTE and/or TruFISP in axial, sagittal, and/or, coronal planes (field of view [FOV] = 380, matrix = 256 × 256, TR/TE = 1350/91 for HASTE and 3.9/1.6 for TruFISP, slice thickness = 3 mm, interslice gap = 0.4 mm). DWI sequence was obtained with b values of 50, 400, and 800, with ADC maps (FOV = 320, matrix = 192 × 100, TR/TE = 7000/101, slice thickness = 6.0 mm, interslice gap = 1.0 mm) in the axial, sagittal, and/or coronal planes with respect to the uterus.

Image analysis

Two sets of DVDs were generated, one with anonymized T2W images and the other with DWI and ADC images of the 42 patients. Two radiologists with 9 and 4 years of experience in prenatal MRI interpreted the images. While the first radiologist was given the set of DVDs with T2W images, the second radiologist reviewed the set with DW images. After completion of interpretation, the DVDs were exchanged between the two radiologists. Both the radiologists were blinded to the clinical, ultrasound, and pathological/surgical data.

The radiologists were given a simple questionnaire of six questions. They were asked to look for the presence of placental heterogeneity, intraplacental dark bands, uterine bulge, loss of placental–myometrial interface, thinning of the myometrium (<2 mm), and any evidence of extrauterine placental extension. Loss of hypointense urinary bladder outline or bladder tenting was considered an extrauterine placental extension. Answers had to be given only as present or absent. The same questionnaire was applied to both T2W as well as DWI interpretations.

After receiving the answers from both the radiologists, a T2W score and a DWI score were assigned to each patient based on clinical, ultrasound, and MRI data. A score of 1 was awarded for the presence of each of the following features: history of previous cesarean delivery; placenta previa; more than three lacunae on ultrasound; marked placental heterogeneity; intraplacental dark bands; uterine bulge; loss of placental–myometrial interface; and myometrial thinning. A score of 2 was awarded for an extrauterine placental extension. The minimum score could be 0, and the maximum score could be 10. The values of the first three of the above parameters were constant for both T2W and DWI scores. The values of the remaining six parameters varied between T2W and DWI scores based on answers to the questionnaire received from the radiologists (Table 1). Based on the scores, patients were graded as low risk (score 0–3), intermediate risk (score 4–6), and high risk (score 7–10) for placenta accreta.

Table 1.

Calculation of T2W and DWI scoring system used in the study.

		Maximum score
Clinical risk feature	History of previous cesarean delivery	1
Ultrasound features	Placenta previa	1
Ultrasound features	>3 placental lacunae	1
MRI features		T2W	DWI
	Marked placental heterogeneity	1	1
	Intraplacental dark bands	1	1
	Uterine bulge	1	1
	Loss of placental myometrial interface	1	1
	Myometrial thinning	1	1
	Extrauterine placental extension	2	2
Maximum total score		10	10

DWI, diffusion-weighted imaging; MRI, magnetic resonance imaging; T2W, T2-weighted.

Statistical analysis

Frequencies and percentages of clinical data and various imaging features were calculated. The sensitivity, specificity, positive and negative predictive values of all individual MRI parameters were calculated for T2W and DWI interpretations. For overall performance, a score of >6 (high risk) was taken as test positive for placenta accreta for both T2W and DWI. Chi-square tests were used to determine any statistical differences between T2W and DWI interpretations. Inter-reader agreement between the two radiologists for T2W and DWI scores was calculated using Cohen's kappa coefficient. SPSS version 28 (IBM Corp., Armonk, NY, USA) was used for all statistical analysis. A P value <0.05 was considered significant.

Table 2.

Summary of MRI findings in the placenta accreta spectrum.

Finding	Description
Placental heterogeneity	Widely different signals in different parts of the placenta
Intraplacental dark bands	Linear dark lines within the placenta
Uterine bulge	Abnormal convexity of the outer contour of the uterus overlying the placenta
Indistinct placental–myometrial interface	Inability to delineate the placenta and myometrium separately
Myometrial thinning	Thickness of the myometrium <2 mm overlying the placenta
Extrauterine placental extension	Loss of India ink artifact of the posterosuperior urinary bladder wall

MRI, magnetic resonance imaging.

Results

Out of the 42 women in the study, 10 were confirmed to have placenta accreta surgically or pathologically. The mean age of patients with placenta accreta was 32.4 ± 4.1 years, and the mean age of patients without placenta accreta was 28.2 ± 3.2 years. Of the 42 patients, 21 (50%) had a history of previous cesarean delivery, and placenta previa was seen in 17 (40.5%) patients. More than three placental lacunae on ultrasound were seen in 11 of the 42 (26.2%) patients. All the clinical, ultrasound, and MRI parameters included in our scoring system were significantly associated with placenta accreta on pathology or surgery (Table 3). As per T2W scores, 29 patients were placed in the low-risk, four in the intermediate-risk, and nine in the high-risk categories. On DWI interpretations, 30 patients were placed in the low-risk, four in the intermediate-risk, and eight in the high-risk category.

Table 3.

Associations between imaging findings and diagnosis of placenta accreta on surgery/pathology.

Features	Placenta accreta present (n = 10)	Placenta accreta absent (n = 32)	P value
History of previous cesarean delivery	9 (90)	12 (37.5)	0.009
Placenta previa	10 (100)	7 (21.9)	<0.001
>3 lacunae on USG	7 (70)	4 (12.5)	0.001
T2W interpretation
Placental heterogeneity	9 (90)	7 (21.9)	<0.001
Intraplacental dark bands	10 (100)	6 (18.8)	<0.001
Uterine bulge	6 (60)	2 (6.3)	<0.001
Loss of placental–uterine interface	9 (90)	7 (21.9)	<0.001
Myometrial thickness <2 mm	10 (100)	6 (18.8)	<0.001
Extrauterine placental extension	5 (50)	0 (0)	<0.001
Low risk (T2W score 0–3)	0 (0)	29 (90.6)	<0.001
Intermediate risk (T2W score 4–6)	1 (10)	3 (9.4)
High risk (T2W score 7–10)	9 (90)	0 (0)
DWI interpretation
Placental heterogeneity	10 (100)	9 (28.1)	<0.001
Intraplacental dark bands	10 (100)	9 (28.1)	0.001
Uterine bulge	6 (60)	2 (6.3)	<0.001
Loss of placental–uterine interface	9 (90)	1 (3.1)	<0.001
Myometrial thickness <2 mm	9 (90)	2 (6.3)	<0.001
Extrauterine placental extension	6 (60)	0 (0)	<0.001
Low risk (DWI score 0–3)	0 (0)	30 (93.8)	<0.001
Intermediate risk (DWI score 4–6)	2 (20)	2 (6.3)
High risk (DWI score 7–10)	8 (80)	0 (0)

Values are given as n (%).

DWI, diffusion-weighted imaging; MRI, magnetic resonance imaging; T2W, T2-weighted; USG, ultrasonography.

In cases with placental invasion, the placenta appeared more heterogeneous, i.e it had areas of widely different signal in different parts. In addition, intraplacental dark bands were seen in these cases as linear dark lines within the placental tissue. A uterine bulge was seen as an additional feature as abnormal convexity of the outer contour of uterus in regions overlying the placenta. An attempt was made to differentiate the placenta from the underlying myometrium based on their contrast difference. If the underlying myometrium could not be differentiated from the placenta, the placental–myometrial interface was labeled as indistinct. The thickness of the myometrium was measured, in the region overlying the placenta. If it was <2 mm, it was labeled a thinned myometrium. Finally, extrauterine placental invasion was assessed best on sagittal images by identifying placental tissue penetrating through the uterine serosa, bladder wall tenting, and loss of India ink artifact of the posterosuperior urinary bladder wall (Table 2).

On T2W interpretations, intraplacental dark bands and myometrial thinning were the most sensitive parameter (100% sensitivity), whereas on DWI interpretations, placental heterogeneity and intraplacental dark bands were the most sensitive parameter (100% sensitivity). Extrauterine placental extension was the least sensitive parameter on T2W imaging (50%), whereas uterine bulge was the least sensitive parameter for DWI interpretations (60% sensitivity). An extrauterine placental extension was the most specific parameter (100% specificity) and had the highest positive predictive value (100% PPV) for both T2W and DWI interpretations. Intraplacental dark bands and myometrial thinning on T2W and placental heterogeneity and intraplacental dark bands on DWI interpretations had the highest negative predictive values.

T2W and DWI interpretations were comparable. There were no statistical differences in individual parameters and T2W and DWI scores between T2W and DWI interpretations. Considering a cutoff score of >6 as test positive, the sensitivity (90% vs. 80%) and negative predictive value (96.9% vs. 94.1%) of T2W score were higher than that of DWI score. The specificity and positive predictive value of both T2W and DWI scores were 100% (Table 4).

Table 4.

Diagnostic accuracy of T2W and DWI interpretations.

MRI features	T2W interpretation				DWI interpretation				P value
MRI features	Sensitivity (%)	Specificity (%)	PPV (%)	NPV (%)	Sensitivity (%)	Specificity (%)	PPV (%)	NPV (%)	P value
Placental heterogeneity	90	78.1	56.3	96.2	100	71.9	52.6	100	0.250
Intraplacental dark bands	100	81.3	62.5	100	100	71.9	52.6	100	0.375
Uterine bulge	60	93.8	75	85.2	60	93.8	75	88.2	1.0
Loss of placental–myometrial interface	90	90.6	75	96.7	90	96.9	90	96.9	0.5
Myometrial thinning	100	81.3	62.5	100	90	93.8	81.8	96.8	0.063
Extrauterine placental extension	50	100	100	86.5	60	100	100	88.9	1.0
Total score >6	90	100	100	96.9	80	100	100	94.1	0.513

DWI, diffusion-weighted imaging; MRI, magnetic resonance imaging; T2W, T2-weighted.

There was substantial inter-reader agreement between the T2W and DWI scores of the two radiologists. Cohen's kappa coefficient of inter-reader agreement between the two radiologists was higher for T2W scores than DWI scores (0.943 vs. 0.882).

Discussion

Our study's results of T2W and DWI interpretations were comparable. The T2W and DWI scores used by us were based on a combination of clinical, ultrasound, and MRI parameters. Though the sensitivity and negative predictive value of T2W scores were higher than DWI scores, there were no significant statistical differences between the two. T2W scores also had higher inter-reader agreement than DWI scores.

We conceived the idea to compare T2W and DWI sequences for placenta accreta due to a few recent studies that demonstrated that DWI sequences showed an intrinsic higher signal of the placenta due to high placental cellularity compared to the low signal of the myometrium (12) and thus the placental–myometrial interface is better visualized on DWI (2). Most institutes do not include a DWI sequence in their prenatal MRI protocol; thus, experience in interpreting DWI in placenta accreta is limited.

The various features of placenta accreta seen on ultrasound include loss of retroplacental halo, placental lacunae, indistinct endometrial–myometrial interface, and hypervascularization of subplacental zone on Doppler (11). According to Finberg's criteria, lacunae on ultrasound are classified into four grades (grade 0–3). Grade 0 indicates no placental lacunae; grade 1+ indicates 1–3 small lacunae; grade 2+ indicates 4–6 large or irregular lacunae; and grade 3+ indicates multiple lacunae throughout the placenta (14).

Yang et al. studied the association of placental lacunae on ultrasound with placenta accreta in patients with complete placenta previa and a history of prior cesarean delivery. They found that a grade of ≥2+ had sensitivity, specificity, positive and negative predictive values of 100%, 97.2%, 93.8%, and 100%, respectively (15). We also included the presence of >3 lacunae as one of the parameters in our scoring system and found it to be significantly associated with the presence of placenta accreta in pathology. We did not include other ultrasound features of placenta accreta, such as loss of retroplacental halo and indistinct placental–myometrial interface, in our scoring system because the MRI correlates of these ultrasound findings were already included in our score.

The various MRI features of placenta accreta include marked placental heterogeneity, T2 dark bands, uterine bulge, thinning of myometrium, loss of placental–myometrial interface, urinary bladder tenting, and extrauterine placental extension (16,17) (Fig 2). Intraplacental dark bands were 100% sensitive on both T2W and DWI interpretations. Previous studies have also found that intraplacental dark bands are the most common feature of placenta accreta on MRI (3,18). Placental heterogeneity also had a sensitivity of 100% on DWI interpretations in diagnosing placenta accreta. An extrauterine placental extension is the most specific MRI feature. Loss of chemical shift artifact of hypointense outline of posterior urinary bladder and bladder tenting signify extrauterine placental invasion of the urinary bladder wall (19). Therefore, to increase the accuracy of our scoring system, a score of 2 was given to this parameter compared to a score of 1 for other parameters.

Similar to previous studies that found DWI better in visualizing the placental–myometrial interface, the present study also found that this parameter had higher specificity and positive and negative predictive values on DWI than on T2W interpretations, though the sensitivity was the same (Fig 3). Similarly, myometrial thinning had higher specificity and positive predictive value on DWI interpretations than on T2W interpretations.

Sannananja et al. compared T2W images with T2W + DWI images and found that adding a DWI sequence to a T2W sequence did not bring any significant additional benefit in diagnosing placenta accreta (2). However, the present study is the first study directly comparing T2W and DWI sequences in placenta accreta.

Different scoring systems for placenta accreta have been used previously (20 –23). Their limitations are discussed in Table 5. Our scoring method is more comprehensive as it includes a combination of clinical risk factors, ultrasound, and MRI features. Thus, the sensitivity and specificity of our scoring system are higher than most of the previous scoring methods. Additionally, our scoring system is applicable to both T2W and DWI sequences.

Table 5.

Summary of various scoring systems used by different authors for placenta accreta.

Author	Based on	Sensitivity (%)	Specificity (%)	Limitations
Tovbin et al. (21)	USG findings	69.6	98.7	Clinical and MRI features not included
Ueno et al. (22)	MRI findings	83.3	94.2	Clinical and ultrasound features not included Each finding had multiple possible scores leading to subjectivity
Knight et al. (22)	USG and MRI	56	92	Clinical risk factors not included Low sensitivity
Tanimura et al. (22)	Clinical, USG, and MRI	91.3	98	Included only patients with placenta previa

MRI, magnetic resonance imaging; T2W, T2-weighted; USG, ultrasonography.

Our study has various clinical implications. First, it gives an insight into the use of DWI in prenatal MRI and placenta accreta. Since there are no statistical differences between T2W and DWI interpretations in placenta accreta, one may safely use either of these sequences without any significant risk of missing or overdiagnosing any imaging feature. Often, motion artifacts degrade T2W images in prenatal MRI. As DW images are resistant to motion, they may alternatively help in visualizing areas and interfaces that are blurred on T2W sequence without compromising the accuracy of the result (Fig 4). DWI has a lower resolution as compared to T2W images. This could be the reason for the higher sensitivity of T2W interpretations than DWI interpretations in our study.

The present study has some limitations. It was a retrospective study with relatively small sample size. We did not differentiate between placenta accreta, increta, and percreta and compare the performance of T2W and DWI interpretations in determining the depth of placental invasion. Both radiologists evaluated the set of T2W images and DWI images one after the other. This could have led to some recall bias.

In conclusion, T2W and DWI sequences are comparable in evaluating the placenta accreta spectrum. A comprehensive scoring system based on clinical, ultrasound, and MRI features has a high accuracy for diagnosing placenta accreta. T2W scores have higher sensitivity, negative predictive value, and inter-reader agreement than DWI scores, although there are no significant differences in the overall performance of T2W and DWI sequences.

Footnotes

Acknowledgements

The authors acknowledge their team of MR technicians led by Mrs. Rachel Luyees for their role in this study.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article

ORCID iDs

Adity Prakash

Ishan Kumar

Ashish Verma

References

Mazouni

Gorincour

Juhan

, et al. Placenta accreta: A review of current advances in prenatal diagnosis. Placenta 2007;28:599–603.

Sannananja

Ellermeier

Hippe

, et al. Utility of diffusion-weighted MR imaging in the diagnosis of placenta accreta spectrum abnormality. Abdom Radiol (NY) 2018;43:3147–3156.

Derman

Nikac

Haberman

, et al. MRI of placenta accreta: A new imaging perspective. AJR Am J Roent 2011;197:1514–1521.

Bretelle

Courbiere

Mazouni

, et al. Management of placenta accreta: Morbidity and outcome. Eur J Obstet Gynecol Reprod Biol 2007;133:34–39.

Hull

Moore

. Multiple repeat cesareans and the threat of placenta accreta: Incidence, diagnosis, management. ClinPerinatol 2011;38:285–296.

Kocherginsky

Hibbard

. Abnormal placentation: Twenty-year analysis. Am J Obstet Gynecol 2005;192:1458–1461.

Silver

. The MFMU cesarean section registry: Maternal morbidity associated with multiple repeat cesarean delivery. Am J Obstet Gynecol 2004;191:S17.

Clark

Koonings

Phelan

. Placenta previa/accreta and prior cesarean section. Obstet Gynecol 1985;66:89–92.

Baughman

Corteville

Shah

. Placenta accreta: Spectrum of US and MR imaging findings. Radiographics 2008;28:1905–1916.

10.

Jauniaux

Bhide

. Prenatal ultrasound diagnosis and outcome of placenta previa accreta after cesarean delivery: A systematic review and meta-analysis. Am J Obstet Gynecol 2017;217:27–36.

11.

D’Antonio

Iacovella

Bhide

. Prenatal identification of invasive placentation using ultrasound: Systematic review and meta-analysis. Ultrasound Obstet Gynecol 2013;42:509–517.

12.

Morita

Ueno

Fujimura

, et al. Feasibility of diffusion-weighted MRI for defining placental invasion. J Magn Reson Imaging 2009;30:666–671.

13.

Konijeti

Rajfer

Askari

. Placenta percreta and the urologist Rev Urol 2009;11:173–176.

14.

Finberg

Williams

. Placenta accreta: Prospective sono- graphic diagnosis in patients with placenta previa and prior cesarean section. J Ultrasound Med 1992;11:333–343.

15.

Yang

Lim

Kim

, et al. Sonographic findings of placental lacunae and the prediction of adherent placenta in women with placenta previa totalis and prior Cesarean section. Ultrasound Obstet Gynecol 2006;28:178–182.

16.

Kumar I, Verma A, Ojha R, Shukla RC, Jain M, Srivastava A. Invasive placental disorders: a prospective US and MRI comparative analysis. Acta Radiologica. 2017;58(1):121–128

17.

Kumar I, Verma A, Jain M, Shukla RC. Structured evaluation and reporting in imaging of placenta and umbilical cord. Acta Radiologica. 2020;61(5):685–704.

18.

Lax

Prince

Mennitt

, et al. The value of specific MRI features in the evaluation of suspected placental invasion. Magn Reson Imaging 2007;25:87–93.

19.

Kumar I, Verma A, Jain S, Jain M, Shukla RC, Srivastava A. Chemical Shift Artifact on Steady-State MRI Sequences for Detection of Vesical Wall Invasion in Placenta Percreta. J Obstet Gynaecol India. 2016;66(2):101–106

20.

Tovbin

Melcer

Shor

, et al. Prediction of morbidly adherent placenta using a scoring system. Ultrasound Obstet Gynecol 2016;48:504–510.

21.

Ueno

Maeda

Tanaka

, et al. Evaluation of interobserver variability and diagnostic performance of developed MRI-based radiological scoring system for invasive placenta previa. J Magn Reson Imaging 2016;44:573–583.

22.

Knight

Lehnert

Shanks

, et al. A comprehensive severity score for the morbidly adherent placenta: Combining ultrasound and magnetic resonance imaging. Pediatr Radiol 2018;48:1945–1954.

23.

Tanimura

Morizane

Deguchi

, et al. A novel scoring system for predicting adherent placenta in women with placenta previa. Placenta 2018;64:27–33.