Accuracy of MRI-Based Radiomics in Diagnosis of Placenta Accreta Spectrum: A PRISMA Systematic Review and Meta-Analysis.

BACKGROUND Placenta accreta syndrome (PAS) can lead to severe obstetric bleeding, and can be life-threatening. This study aimed to assess the precision of radiomics features derived from magnetic resonance imaging (MRI) for diagnosing PAS. MATERIAL AND METHODS A comprehensive search was conducted in the databases PubMed, Embase, Web of Science, and the Cochrane library from inception to October 2023. We included diagnostic accuracy studies utilizing radiomics-MRI in PAS patients, with histopathology serving as the reference standard. The overall diagnostic odds ratio (DOR), sensitivity, specificity, and area under the curve (AUC) were computed to gauge the diagnostic accuracy of MRI-based radiomic features in PAS patients. Quality assessment was performed using the Quality Assessment of Diagnostic Accuracy Studies 2. Statistical analyses were carried out using Stata 14.2, MetaDiSc 1.4, and Review Manager 5.3 software. RESULTS Seven studies involving 672 patients were incorporated. The aggregated DOR, sensitivity, specificity, and AUC for radiomics in detecting PAS were 78% (confidence interval32, 191), 87% (76%, 93%), 92% (89%, 94%), and 0.93 (0.91-0.95), respectively. The meta-analysis revealed notable heterogeneity among the included studies, with no evidence of a threshold effect. Subgroup analysis demonstrated that, in comparison to manual segmentation and validation groups with ≤100 cases and internal validation datasets, automated segmentation, validation groups with >100 cases, and external validation datasets exhibited superior diagnostic performance . CONCLUSIONS Our findings indicate that MRI-based radiomic features perform well in assessing the diagnostic risk of PAS during prenatal diagnosis. This noninvasive and convenient tool may prove valuable in facilitating the identification of PAS.

Diffusion-derived vessel density (DDVD) computed from a simple diffusion MRI protocol as a biomarker of placental blood circulation in patients with placenta accreta spectrum disorders: A proof-of-concept study.

OBJECTIVES: Increasing trend of PAS (placenta accreta spectrum disorders) incidence is a major health concern as PAS is associated with high maternal morbidity and mortality during cesarean section. Prenatal identification of PAS is crucial for delivery planning and patients management. This study aims to explore whether diffusion-derived vessel density (DDVD) computed from a simple diffusion MRI protocol differs in PAS from normal placenta. METHODS: We enrolled 86 patients with PAS disorders and 40 pregnant women without PAS disorders. Each patient underwent intravoxel incoherent motion (IVIM) MRI sequence with 11 b-values. Placenta diffusion-derived vessel density (DDVD-b0b50) was the signal difference between b = 0 and b = 50 s/mm2 images. DDVD(b0b50) A/N was calculated as [accreta lesion DDVD(b0b50)]/ [normal placenta DDVD(b0b50)]. The correlation between DDVD and gestational age was explored using Spearman rank correlation. Differences of DDVD(b0b50) A/N in patients with normal placentas and with PAS, and in patients with different subtypes of PAS were explored. RESULTS: DDVD was negatively correlated with gestational age (p = 0.023, r = -0.359) in patients with normal placentas. DDVD(b0b50) A/N was significantly higher in patients with PAS (median:1.16, mean: 1.261) than normal placenta (median:1.02, mean: 1.032, p < 0.001) and especially higher in patients with placenta increta (median:1.14, mean: 1.278) and percreta (median: 1.20, mean: 1.396, p < 0.001). CONCLUSION: As a higher DDVD indicates higher physiological volume of micro-vessels in PAS, this study suggests DDVD can be a potential biomarker to evaluate the placenta perfusion.

The value of the combined MR imaging features and clinical factors Nomogram model in predicting intractable postpartum hemorrhage due to placenta accreta.

To explore the value of the combined MR imaging features and clinical factors Nomogram model in predicting intractable postpartum hemorrhage (IPH) due to placenta accreta (PA). We conducted a retrospective study with 270 cases of PA patients admitted to our hospital from January 2015 to December 2022. The clinical data of these patients were analyzed, and they were divided into 2 groups: the IPH group and the non-IPH group based on the presence of IPH. The differences in data between the 2 groups were compared, and the risk factors for IPH were analyzed. A Nomogram model was constructed using independent high-risk factors, and the predictive value of this model for IPH was analyzed. The results of multivariable binary Logistic regression analysis showed higher number of cesareans, placenta previa, placenta accreta type (implantation, penetration), low signal strip on T2 weighted image (T2WI) were independent high-risk factor for IPH (P < .05). ROC analysis and Hosmer-Lemeshow goodness-of-fit test showed the Nomogram predictive model constructed with the high-risk factor has good discrimination and calibration. Decision curve analysis (DCA) showed that when the probability threshold for the Nomogram model's prediction was in the range from 0.125 to 0.99, IPH patients could obtain more net benefits, making it suitable for clinical application. The higher number of cesareans, placenta previa, placental accreta type (implantation, penetration), and low signal strip on T2WI are independent high-risk factor for IPH. The Nomogram predictive model constructed with the high-risk factor demonstrates good clinical efficacy in predicting the occurrence of IPH due to PA.

Three-dimensional volume rendering ultrasound for assessing placenta accreta spectrum severity and discriminating it from simple scar dehiscence.

BACKGROUND: Prenatal ultrasound discrimination between placenta accreta spectrum and scar dehiscence with underlying nonadherent placenta is challenging both prenatally and intraoperatively, which often leads to overtreatment. In addition, accurate prenatal prediction of surgical difficulty and morbidity in placenta accreta spectrum is difficult, which precludes appropriate multidisciplinary planning. The advent of advanced 3-dimensional volume rendering and contrast enhancement techniques in modern ultrasound systems provides a comprehensive prenatal assessment, revealing details that are not discernible in traditional 2-dimensional imaging. OBJECTIVE: This study aimed to evaluate the use of 3-dimensional volume rendering ultrasound techniques in determining the severity of placenta accreta spectrum and distinguishing between placenta accreta spectrum and scar dehiscence with underlying nonadherent placenta. STUDY DESIGN: A prospective, cohort study was conducted between July 2022 and July 2023 in the fetal medicine unit of Dr Soetomo Academic General Hospital, Surabaya, Indonesia. All pregnant individuals with anterior low-lying placenta or placenta previa with a previous caesarean section who were referred with suspicion of placenta accreta spectrum were consented and screened using the standardised 2-dimensional and Doppler ultrasound imaging. Additional 3-dimensional volumes were obtained from the sagittal section of the uterus with a filled urinary bladder. These were analyzed by rotating the region of interest to be perpendicular to the uterovesical interface. The primary outcomes were the clinical and histologic severity in the cases of placenta accreta spectrum and correct diagnosis of dehiscence with nonadherent placenta underneath. The strength of association between ultrasound and clinical outcomes was determined. Multivariate logistic regression analyses and diagnostic testing of accuracy were used to analyze the data. RESULTS: A total of 70 patients (56 with placenta accreta spectrum and 14 with scar dehiscence) were included in the analysis. Multivariate logistic regression of all 2-dimensional and 3-dimensional signs revealed the 3-dimensional loss of clear zone (P<.001) and the presence of bridging vessels on 2-dimensional Doppler ultrasound (P=.027) as excellent predictors in differentiating scar dehiscence and placenta accreta spectrum. The 3-dimensional loss of clear zone demonstrated a high diagnostic accuracy with an area under the curve of 0.911 (95% confidence interval, 0.819-1.002), with a sensitivity of 89.3% (95% confidence interval, 78.1-95.97%) and specificity of 92.9% (95% confidence interval, 66.1-99.8%). The presence of bridging vessels on 2-dimensional Doppler demonstrated an area under the curve of 0.848 (95% confidence interval, 0.714-0.982) with a sensitivity of 91.1% (95% confidence interval, 80.4-97.0%) and specificity of 78.6% (95% confidence interval, 49.2-95.3%). A subgroup analysis among the placenta accreta spectrum group revealed that the presence of a 3-dimensional disrupted bladder serosa with obliteration of the vesicouterine space was associated with vesicouterine adherence (P<.001). CONCLUSION: Three-dimensional volume rendering ultrasound is a promising tool for effective discrimination between scar dehiscence with underlying nonadherent placenta and placenta accreta spectrum. It also shows potential in predicting the clinical severity with urinary bladder involvement in cases of placenta accreta spectrum.

A novel MRI-based diagnostic model for predicting placenta accreta spectrum.

Objective To develop and evaluate a diagnostic model based on MRI signs for predicting placenta accreta spectrum. Materials and Methods A total of 155 pregnant women were included in this study, randomly divided into 104 cases in the training set and 51 cases in the validation set. There were 93 Non-PAS cases, and 62 cases in the PAS group. The training set included 62 Non-PAS cases and 42 PAS cases. Clinical factors and MRI signs were collected for univariate analysis. Then, binary logistic regression analysis was used to develop independent diagnostic models with clinical relevant risk factors or MRI signs, as well as those combining clinical risk factors and MRI signs. The ROC curve analysis was used to evaluate the diagnostic performance of each diagnostic model. Finally, the validation was performed with the validation set. Results In the training set, four clinical factors (gestity, parity, uterine surgery history, placental position) and 11 MRI features (T2-dark bands, placental bulge, T2 hypointense interface loss, myometrial thinning, bladder wall interruption, focal exophytic mass, abnormal placental bed vascularization, placental heterogeneity, asymmetric placental thickening/shape, placental ischemic infarction, abnormal intraplacental vascularity) were considered as risk factors for PAS. The AUC of the clinical diagnostic model, MRI diagnostic model, and clinical + MRI model of PAS were 0.779, 0.854, and 0.874, respectively. In the validation set, the AUC of the clinical diagnostic model, MRI diagnostic model, and clinical + MRI model of PAS were 0.655, 0.728, and 0.735, respectively. Conclusion Diagnosis model based on MRI features in this study can well predict placenta accreta spectrum.

Sensitivity of antenatal ultrasound in diagnosing posterior placenta accreta spectrum disorders.

OBJECTIVES: Optimal management of placenta accreta spectrum (PAS) requires antenatal diagnosis. We sought to evaluate the sensitivity of ultrasound findings suggestive of PAS in detecting posterior PAS. METHODS: Cohort study of patients with posterior placentation and pathology-confirmed PAS from 2011 to 2020 at a tertiary center. Patients were excluded if ultrasound images were unavailable. Ultrasounds were reviewed for presence of lacunae, hypervascularity, myometrial thinning, loss of the hypoechoic zone, bridging vessels, abnormal uterine serosa-bladder interface, placental bulge, placental extension into/beyond the myometrium, and an exophytic mass. Risk factors, postpartum outcomes, and ultrasound findings were compared by antepartum suspicion for PAS. Sensitivity was calculated for each ultrasound finding. RESULTS: Thirty-three patients were included. PAS was not suspected antenatally in 70 % (23/33). Patients with unsuspected PAS were more likely to be non-Hispanic, have in vitro fertilization, no prior Cesarean deliveries, no placenta previa, and delivered later in gestation. Depth of invasion and estimated blood loss were less for unsuspected PAS, but there was no difference in hysterectomy between groups. Ultrasound findings were less frequently seen in those who were not suspected antenatally: lacunae 17.4 vs. 100 % (p<0.001), hypervascularity 8.7 vs. 80 % (p<0.001), myometrial thinning 4.4 vs. 70 % (p<0.001), and placental bridging vessels 0 vs. 60 % (p<0.001). There was poor sensitivity (0-42.4 %) for all findings. CONCLUSIONS: Posterior PAS is less likely to be detected antenatally due to a lower sensitivity of typical ultrasound findings in the setting of a posterior placenta. Further studies are needed to better identify reliable markers of posterior PAS.

Magnetic resonance imaging improves diagnosis of placenta accreta spectrum requiring hysterectomy compared to ultrasound.

BACKGROUND: Magnetic resonance imaging has been used increasingly as an adjunct for ultrasound imaging for placenta accreta spectrum assessment and preoperative surgical planning, but its value has not been established yet. The ultrasound-based placenta accreta index is a well-validated standardized approach for placenta accreta spectrum evaluation. Placenta accreta spectrum-magnetic resonance imaging markers have been outlined in a joint guideline from the Society of Abdominal Radiology and the European Society of Urogenital Radiology. OBJECTIVE: This study aimed to compare placenta accreta spectrum-magnetic resonance imaging parameters with the ultrasound-based placenta accreta index in pregnancies at high risk for placenta accreta spectrum and to assess the additional diagnostic value of magnetic resonance imaging for placenta accreta spectrum that requires a cesarean hysterectomy. STUDY DESIGN: This was a single-center, retrospective study of pregnant patients who underwent magnetic resonance imaging, in addition to ultrasonography, because of suspected placenta accreta spectrum. The ultrasound-based placenta accreta index and placenta accreta spectrum-magnetic resonance imaging parameters were obtained. Student's t test and Fisher's exact test were used to compare the groups in terms of the primary outcome (hysterectomy vs no hysterectomy). The diagnostic performance of magnetic resonance imaging and the ultrasound-based placenta accreta index was assessed using multivariable logistic regressions, receiver operating characteristics curves, the DeLong test, McNemar test, and the relative predictive value test. RESULTS: A total of 82 patients were included in the study, 41 of whom required a hysterectomy. All patients who underwent a hysterectomy met the International Federation of Gynecology and Obstetrics clinical evidence of placenta accreta spectrum at the time of delivery. Multiple parameters of the ultrasound-based placenta accreta index and placenta accreta spectrum-magnetic resonance imaging were able to predict hysterectomy, and the parameter of greatest dimension of invasion by magnetic resonance imaging was the best quantitative predictor. At 96% sensitivity for hysterectomy, the cutoff values were 3.5 for the ultrasound-based placenta accreta index and 2.5 cm for the greatest dimension of invasion by magnetic resonance imaging. Using this sensitivity, the parameter of greatest dimension of invasion measured by magnetic resonance imaging had higher specificity (P=.0016) and a higher positive predictive value (P=.0018) than the ultrasound-based placenta accreta index, indicating an improved diagnostic threshold. CONCLUSION: In a suspected high-risk group for placenta accreta spectrum, magnetic resonance imaging identified more patients who will not need a hysterectomy than when using the ultrasound-based placenta accrete index only. Magnetic resonance imaging has the potential to aid patient counseling, surgical planning, and delivery timing, including preterm delivery decisions for patients with placenta accreta spectrum requiring hysterectomy.

Application of Uterine Artery Embolization in Patients With Placenta Accreta Spectrum After Abdominal Aortic Balloon Occlusion.

OBJECTIVE: To evaluate the application of different uterine artery embolization procedures under balloon occlusion of the abdominal aorta in patients with Placenta Accreta Spectrum (PAS) undergoing cesarean section. MATERIALS AND METHODS: A retrospective analysis was performed on clinical data from 72 patients who underwent uterine artery embolization for hemostasis during cesarean section with PAS. The patients were divided into two groups according to the embolization method used during surgery: group A (n = 43) underwent uterine artery embolization by withdrawing the balloon and inserting a Cobra catheter into the uterine artery for embolization, while group B (n = 29) underwent uterine artery embolization with a Cobra catheter inserted via contralateral puncture of the femoral artery and balloon occlusion. General information, surgical data, and postoperative recovery were compared between the 2 groups. RESULTS: The bleeding and transfusion volumes were lower in group B than in group A and the differences between the 2 groups were statistically significant. There were no significant differences in surgical duration, number of embolized vessels, length of hospital stay, postoperative complications, or menstrual recovery between the 2 groups. CONCLUSION: For patients with PAS undergoing cesarean section, uterine artery embolization for hemostasis is preferably performed by inserting a Cobra catheter via contralateral puncture of the femoral artery under abdominal aortic balloon occlusion.

Placental volume as a novel sign for identifying placenta accreta spectrum in pregnancies with complete placenta previa.

BACKGROUND: Placenta accreta spectrum (PAS) carries an increased risk of maternal-fetal mortality and morbidity, and magnetic resonance imaging (MRI) features for PAS have been used for preoperative identification. This study aims to investigate the role of placental volume evaluated by MRI in identifying PAS in pregnant women with complete placenta previa. METHODS: Totally 163 cases of complete placenta previa pregnant women with a history of cesarean section underwent MRI for suspected PAS were included. We categorized the patients into two groups according to the presence or absence of PAS, and the maternal-fetal perinatal outcomes and placental volume analyzed by 3D Slice software were compared. RESULTS: There were significantly more gravidity, parity, and number of previous cesarean delivery in the PAS group (P < 0.05). Significant differences were also found between the two groups with respect to the following baseline characteristics: gestational age at delivery, intraoperative blood loss, blood transfusion, and neonatal birth weight (P < 0.05). Of 163 women in the study, 7 (4.294%) required cesarean hysterectomy for high-grade PAS or pernicious bleeding during cesarean section, and PAS was confirmed with histologic confirmation in 6 (85.714%) cases. The placental volume in PAS group was greater than that in the non-PAS group (P < 0.05). With a threshold of more than 887 cm3, the sensitivity and specificity in identifying PAS were 85.531% and 83.907% respectively, with AUC 0.908 (95% CI: 0.853-0.948). CONCLUSIONS: Placental volume may be a promising indicator of PAS in complete placenta previa patients with a history of cesarean section.

The diagnostic efficiency of diffusion-weighted imaging in placenta accreta spectrum: a systematic review and meta-analysis.

OBJECTIVE: This study aims to evaluate the diagnostic efficiency of diffusion-weighted imaging (DWI) in patients with placenta accreta spectrum (PAS). MATERIALS AND METHODS: The present study searched on PubMed, Embase, OVID, Cochrane, Scopus and CNKI, Chinese Bio-Medical Literature, VIP, Wanfang, Duxiu, databases for studies related to the diagnostic performance of DWI for PAS from inception to December 2022. The pooled sensitivity, the pooled specificity, positive likelihood ratio (LR+), negative likelihood ratio (LR-), and diagnosis odds ratios (DOR) were calculated by Meta-disc 1.4 and STATA 16.0. RESULTS: A total of 11 studies met the criteria and were included in the meta-analysis. The effect indexes of DWI in combined PAS were as follows. The pooled sensitivity was 0.670 (0.619-0.719). The pooled specificity was 0.720 (0.661-0.773). The pooled LR+ was 2.161 (1.454-3.211). The pooled LR- was 0.413 (0.280-0.609). The pooled AUC was 0.7841, and Q* was 0.7221. The pooled diagnostic ratio DOR was 6.713 (2.981-15.118). Subgroup analysis showed that four studies used T2-weighted imaging (T2WI) + DWI to diagnose PAS, and the pooled AUC was 0.9822. CONCLUSIONS: The results showed that DWI had high sensitivity and specificity in the diagnosis of PAS. Furthermore, T2WI+DWI has higher diagnostic efficacy than DWI alone in the diagnosis of PAS. Therefore, it is necessary to set T2WI+DWI as a routine sequence for PAS, and T2WI+DWI should be a routine method for the daily diagnosis of PAS.

Diagnostic Role of Cell-Free miRNAs in Identifying Placenta Accreta Spectrum during First-Trimester Screening.

Placenta accreta spectrum (PAS) is a severe complication of pregnancy associated with excessive invasion of cytotrophoblast cells at the sites of the endometrial-myometrial interface and the myometrium itself in cases of adherent (creta) and invasive (increta and percreta) forms, respectively. This leads to a high risk of massive blood loss, maternal hysterectomy, and preterm birth. Despite advancements in ultrasound protocols and found associations of alpha-fetoprotein, PAPP-A, hCG, PLGF, sFlt-1, IL-8, and IL-33 peripheral blood levels with PAS, there is a high need for an additional non-invasive test to improve the diagnostic accuracy and to select the real PAS from the suspected ones in the first-trimester screening. miRNA signatures of placental tissue, myometrium, and blood plasma from women with PAS in the third trimester of pregnancy, as well as miRNA profiles in exosomes from the blood serum of women in the first trimester with physiologically progressing pregnancy, complicated by PAS or pre-eclampsia, were obtained using deep sequencing. Two logistic regression models were constructed, both featuring statistically significant parameters related to the levels of miR-26a-5p, miR-17-5p, and miR-101-3p, quantified by real-time PCR in native blood serum. These models demonstrated 100% sensitivity in detecting PAS during the first pregnancy screening. These miRNAs were identified as specific markers for PAS, showing significant differences in their blood serum levels during the first trimester in the PAS group compared to those in physiological pregnancies, early- or late-onset pre-eclampsia groups. Furthermore, these miRNAs exhibited differential expression in the PAS placenta and/or myometrium in the third trimester and, according to data from the literature, control angiogenesis. Significant correlations were found between extracellular hsa-miR-101-3p and nuchal translucency thickness, hsa-miR-17-5p and uterine artery pulsatility index, and hsa-miR-26a-5p and hsa-miR-17-5p with PLGF. The developed test system for early non-invasive PAS diagnosis based on the blood serum level of extracellular miR-26a-5p, miR-17-5p, and miR-101-3p can serve as an auxiliary method for first-trimester screening of pregnant women, subject to validation with independent test samples.

Development and validation of MRI-based scoring models for predicting placental invasiveness in high-risk women for placenta accreta spectrum.

OBJECTIVES: To develop and validate MRI-based scoring models for predicting placenta accreta spectrum (PAS) invasiveness. MATERIALS AND METHODS: This retrospective study comprised a derivation cohort and a validation cohort. The derivation cohort came from a systematic review of published studies evaluating the diagnostic performance of MRI signs for PAS and/or placenta percreta in high-risk women. The significant signs were identified and used to develop prediction models for PAS and placenta percreta. Between 2016 and 2021, consecutive high-risk pregnant women for PAS who underwent placental MRI constituted the validation cohort. Two radiologists independently evaluated the MRI signs. The reference standard was intraoperative and pathologic findings. The predictive ability of MRI-based models was evaluated using the area under the curve (AUC). RESULTS: The derivation cohort included 26 studies involving 2568 women and the validation cohort consisted of 294 women with PAS diagnosed in 258 women (88%). Quantitative meta-analysis revealed that T2-dark bands, placental/uterine bulge, loss of T2 hypointense interface, bladder wall interruption, placental heterogeneity, and abnormal intraplacental vascularity were associated with both PAS and placenta percreta, and myometrial thinning and focal exophytic mass were exclusively associated with PAS. The PAS model was validated with an AUC of 0.90 (95% CI: 0.86, 0.93) for predicting PAS and 0.85 (95% CI: 0.79, 0.90) for adverse peripartum outcome; the placenta percreta model showed an AUC of 0.92 (95% CI: 0.86, 0.98) for predicting placenta percreta. CONCLUSION: MRI-based scoring models established based on quantitative meta-analysis can accurately predict PAS, placenta percreta, and adverse peripartum outcome. CLINICAL RELEVANCE STATEMENT: These proposed MRI-based scoring models could help accurately predict PAS invasiveness and provide evidence-based risk stratification in the management of high-risk pregnant women for PAS. KEY POINTS: • Accurately identifying placenta accreta spectrum (PAS) and assessing its invasiveness depending solely on individual MRI signs remained challenging. • MRI-based scoring models, established through quantitative meta-analysis of multiple MRI signs, offered the potential to predict PAS invasiveness in high-risk pregnant women. • These MRI-based models allowed for evidence-based risk stratification in the management of pregnancies suspected of having PAS.

Deep Learning Model Based on Multisequence MRI Images for Assessing Adverse Pregnancy Outcome in Placenta Accreta.

BACKGROUND: Preoperative assessment of adverse outcomes risk in placenta accreta spectrum (PAS) disorders is of high clinical relevance for perioperative management and prognosis. PURPOSE: To investigate the association of preoperative MRI multisequence images and adverse pregnancy outcomes by establishing a deep learning model in patients with PAS. STUDY TYPE: Retrospective. POPULATION: 323 pregnant women (age from 20 to 46, the median age is 33), suspected of PAS, underwent MRI to assess the PAS, divided into the training (N = 227) and validation datasets (N = 96). FIELD STRENGTH/SEQUENCE: 1.5T scanner/fast imaging employing steady-state acquisition sequence and single shot fast spin echo sequence. ASSESSMENT: Different deep learning models (i.e., with single MRI input sequence/two sequences/multisequence) were compared to assess the risk of adverse pregnancy outcomes, which defined as intraoperative bleeding ≥1500 mL and/or hysterectomy. Net reclassification improvement (NRI) was used for quantitative comparison of assessing adverse pregnancy outcome between different models. STATISTICAL TESTS: The AUC, sensitivity, specificity, and accuracy were used for evaluation. The Shapiro-Wilk test and t-test were used. A P value of <0.05 was considered statistically significant. RESULTS: 215 cases were invasive placenta accreta (67.44% of them with adverse outcomes) and 108 cases were non-invasive placenta accreta (9.25% of them with adverse outcomes). The model with four sequences assessed adverse pregnancy outcomes with AUC of 0.8792 (95% CI, 0.8645-0.8939), with ACC of 85.93% (95%, 84.43%-87.43%), with SEN of 86.24% (95% CI, 82.46%-90.02%), and with SPC of 85.62% (95%, 82.00%-89.23%) on the test cohort. The performance of model with four sequences improved above 0.10 comparing with that of model with two sequences and above 0.20 comparing with that of model with single sequence in terms of NRI. DATA CONCLUSION: The proposed model showed good diagnostic performance for assessing adverse pregnancy outcomes. LEVEL OF EVIDENCE: 3 TECHNICAL EFFICACY: Stage 2.

Prenatal Diagnosis of Placenta Accreta Spectrum Disorders: Deep Learning Radiomics of Pelvic MRI.

BACKGROUND: Diagnostic performance of placenta accreta spectrum (PAS) by prenatal MRI is unsatisfactory. Deep learning radiomics (DLR) has the potential to quantify the MRI features of PAS. PURPOSE: To explore whether DLR from MRI can be used to identify pregnancies with PAS. STUDY TYPE: Retrospective. POPULATION: 324 pregnant women (mean age, 33.3 years) suspected PAS (170 training and 72 validation from institution 1, 82 external validation from institution 2) with clinicopathologically proved PAS (206 PAS, 118 non-PAS). FIELD STRENGTH/SEQUENCE: 3-T, turbo spin-echo T2-weighted images. ASSESSMENT: The DLR features were extracted using the MedicalNet. An MRI-based DLR model incorporating DLR signature, clinical model (different clinical characteristics between PAS and non-PAS groups), and MRI morphologic model (radiologists' binary assessment for the PAS diagnosis) was developed. These models were constructed in the training dataset and then validated in the validation datasets. STATISTICAL TESTS: The Student t-test or Mann-Whitney U, χ2 or Fisher exact test, Kappa, dice similarity coefficient, intraclass correlation coefficients, least absolute shrinkage and selection operator logistic regression, multivariate logistic regression, receiver operating characteristic (ROC) curve, DeLong test, net reclassification improvement (NRI) and integrated discrimination improvement (IDI), calibration curve with Hosmer-Lemeshow test, decision curve analysis (DCA). P < 0.05 indicated a significant difference. RESULTS: The MRI-based DLR model had a higher area under the curve than the clinical model in three datasets (0.880 vs. 0.741, 0.861 vs. 0.772, 0.852 vs. 0.675, respectively) or MRI morphologic model in training and independent validation datasets (0.880 vs. 0.760, 0.861, vs. 0.781, respectively). The NRI and IDI were 0.123 and 0.104, respectively. The Hosmer-Lemeshow test had nonsignificant statistics (P = 0.296 to 0.590). The DCA offered a net benefit at any threshold probability. DATA CONCLUSION: An MRI-based DLR model may show better performance in diagnosing PAS than a clinical or MRI morphologic model. LEVEL OF EVIDENCE: 3 TECHNICAL EFFICACY STAGE: 2.

A Deep Learning Pipeline Using Prior Knowledge for Automatic Evaluation of Placenta Accreta Spectrum Disorders With MRI.

BACKGROUND: The diagnosis of prenatal placenta accreta spectrum (PAS) with magnetic resonance imaging (MRI) is highly dependent on radiologists' experience. A deep learning (DL) method using the prior knowledge that PAS-related signs are generally found along the utero-placental borderline (UPB) may help radiologists, especially those with less experience, to mitigate this issue. PURPOSE: To develop a DL tool for antenatal diagnosis of PAS using T2-weighted MR images. STUDY TYPE: Retrospective. SUBJECTS: Five hundred and forty pregnant women with clinically suspected PAS disorders from two institutions, divided into training (409), internal test (103), and external test (28) datasets. FIELD STRENGTH/SEQUENCE: Sagittal T2-weighted fast spin echo sequence at 1.5 T and 3 T. ASSESSMENT: An nnU-Net was trained for placenta segmentation. The UPB straightening approach was used to extract the utero-placental boundary region. The UPB image was then fed into DenseNet-PAS for PAS diagnosis. DenseNet-PP learnt placental position information to improve the PAS diagnosis performance. Three radiologists with 8, 10, and 12 years of experience independently evaluated the images. Two radiologists marked the placenta tissue. Histopathological findings were the reference standard. STATISTICAL TESTS: Area under the curve (AUC) was used to evaluate the classification. Dice coefficient evaluated the segmentation between radiologists and the model performance. The Mann-Whitney U-test or the chi-squared test assessed the significance of differences. Decision curve analysis was used to determine clinical effectiveness. DeLong's test was used to compare AUCs. RESULTS: Of the 540 patients, 170 had PAS disorders confirmed by histopathology. The DL model using UPB images and placental position yielded the highest AUC of 0.860 and 0.897 in internal test and external test cohorts, respectively, significantly exceeding the performance of three radiologists (internal test AUC, 0.737-0.770). DATA CONCLUSION: By extracting the UPB image, this fully automatic DL pipeline achieved high accuracy and may assist radiologists in PAS diagnosis using MRI. LEVEL OF EVIDENCE: 3 TECHNICAL EFFICACY: Stage 2.

Placental lakes vs lacunae: spot the differences.

Sonographic sonolucencies are anechoic areas surrounded by tissue of normal echogenicity, commonly found in the placental parenchyma during the second and third trimesters of pregnancy. The ultrasound appearance of lakes and lacunae derives from the low echogenicity of villous-free areas within the placental parenchyma, filled with maternal blood of varying velocities. In normal placentation, lakes usually start appearing as soon as maternal blood begins to flow freely within the intervillous space at the end of the first trimester, whereas, in accreta placentation, lacunae develop progressively during the second trimester. Larger lakes are found mainly in areas of lower villous density under the fetal plate or in the marginal areas, but can also be found in the center of a lobule above the entry of a spiral artery. Lakes of variable size, position and shape are of no clinical significance, except if they transform into echogenic cystic lesions, which have been associated with poor fetal growth and placental malperfusion. Lacunae are formed by the distortion of one or more placental lobules developing inside a uterine scar, resulting from high-volume, high-velocity flows from the radial/arcuate arteries, and are associated with a high probability of placenta accreta spectrum at birth. They often present with ultrasound signs of uterine remodeling following scarring. Lakes and lacunae can coexist within the same placenta and both will change in size and shape as pregnancy advances. Better understanding of the etiopathology of placental sonolucent spaces and associated morphological changes is necessary to identify patients at risk of subsequent complications during pregnancy and/or at delivery. © 2023 The Authors. Ultrasound in Obstetrics & Gynecology published by John Wiley & Sons Ltd on behalf of International Society of Ultrasound in Obstetrics and Gynecology.