Diagnostic performance of radiologists with different levels of experience in the interpretation of MRI of the placenta accreta spectrum disorder

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.

Segmentation methods applied to MRI-derived radiomic analysis for the prediction of placenta accreta spectrum in patients with placenta previa

PURPOSE

To retrospectively evaluate the performance of different manual segmentation methods of placenta MR images for predicting Placenta Accreta Spectrum (PAS) disorders in patients with placenta previa (PP) using a Machine Learning (ML) Radiomics analysis.

METHODS

64 patients (n=41 with PAS and n= 23 without PAS) with PP who underwent MRI examination for suspicion of PAS were retrospectively selected. All MRI examinations were acquired on a 1.5 T using T2-weighted (T2w) sequences on axial, sagittal and coronal planes. Ten different manual segmentation methods were performed on sagittal placental T2-weighted images obtaining five sets of 2D regions of interest (ROIs) and five sets of 3D volumes of interest (VOIs) from each patient. In detail, ROIs and VOIs were positioned on the following areas: placental tissue, retroplacental myometrium, cervix, placenta with underneath myometrium, placenta with underneath myometrium and cervix. For feature stability testing, the same process was repeated on 30 randomly selected placental MRI examinations by two additional radiologists, working independently and blinded to the original segmentation. Radiomic features were extracted from all available ROIs and VOIs. 100 iterations of 5-fold cross-validation with nested feature selection, based on recursive feature elimination, were subsequently run on each ROI/VOI to identify the best-performing method to classify instances correctly.

RESULTS

Among the segmentation methods, the best performance in predicting PAS was obtained by the VOIs covering the retroplacental myometrium (Mean validation score: 0.761, standard deviation: 0.116).

CONCLUSION

Our preliminary results show that the VOI including the retroplacental myometrium using T2w images seems to be the best method when segmenting images for the development of ML radiomics predictive models to identify PAS in patients with PP.

Diffusion kurtosis and intravoxel incoherent motion in predicting postpartum hemorrhage in patients at high risk for placenta accreta spectrum disorders

Background

Placenta accreta spectrum (PAS) disorder encompasses a spectrum of pathologies, from placenta accreta to placenta percreta, which is usually associated with postpartum hemorrhage (PPH).

Methods

This cross-sectional study enrolled 109 patients suspected of having PAS disorders based on previous ultrasound results or clinical risk factors from November 2018 to March 2022 in Sichuan Provincial People's Hospital. Of the 109 patients, 34 had PPH and 75 did not have PPH. Magnetic resonance imaging (MRI) including diffusion-weighted imaging (DWI), intravoxel incoherent motion (IVIM), and diffusion kurtosis imaging (DKI) was performed for each patient and the apparent diffusion coefficient (ADC) from DWI, perfusion fraction (f), pure diffusion coefficient (D), and pseudo-diffusion coefficient (D*) from IVIM, and mean diffusion kurtosis (MK) and mean diffusion coefficient (MD) from DKI were measured and compared. The correlation between the DWI parameters and estimated blood loss (EBL) during surgery was identified using correlation analysis. The diagnostic performance for predicting PPH was compared between the two methods.

Results

The amount of bleeding during delivery was positively correlated with D [r=0.331, P<0.001, 95% confidence interval (CI): 0.170 to 0.477], D* (r=0.389, P<0.001, 95% CI: 0.207 to 0.527), f (r=0.222, P=0.02, 95% CI: 0.036 to 0.398), and MD (r=0.277, P=0.003, 95% CI: 0.108 to 0.439), but negatively correlated with MK (r=-0.280, P=0.003, 95% CI: -0.431 to -0.098). In predicting PPH, multivariate analyses showed the independent risk factors were placenta previa and D; the area under the curve (AUC) was 0.795 (95% CI: 0.711 to 0.878) when the two risk factors were combined together.

Conclusions

IVIM and DKI parameters are correlated with EBL. The combined use of placenta previa and D are helpful for predicting PPH in patients at high risk of PAS disorders.

Differentiation of placenta percreta through MRI features and diffusion-weighted magnetic resonance imaging

OBJECTIVES

To identify whether parameters measured from diffusion kurtosis and intravoxel incoherent motion help diagnose placenta percreta.

METHODS

We retrospectively enrolled 75 patients with PAS disorders including 13 patients with placenta percreta and 40 patients without PAS disorders. Each patients underwent diffusion-weighted imaging (DWI), intravoxel incoherent motion (IVIM), and diffusion kurtosis imaging (DKI). The apparent diffusion coefficient (ADC), perfusion fraction (f), pure diffusion coefficient (D), pseudo-diffusion coefficient (D*), mean diffusion kurtosis (MK) and mean diffusion coefficient (MD) were measured by the volumetric analysis and compared. MRI features were also analyzed and compared. The receiver operating characteristic (ROC) curve and logistic regression analysis were used to evaluate the diagnostic efficiency of different diffusion parameters and MRI features for distinguishing placental percreta.

RESULTS

D* was an independent risk factor from DWI for predicting placenta percreta with sensitivity of 73% and specificity of 76%. Focal exophytic mass remained as independent risk factor from MRI features for predicting placenta percreta with sensitivity of 72.7% and specificity of 88.1%. When the two risk factors were combined together, the AUC was the highest, 0.880 (95% CI 0.8-0.96).

CONCLUSION

D* and focal exophytic mass were associated with placenta percreta. A combination of the 2 risk factors can be used to predict placenta percreta.

CRITICAL RELEVANCE STATEMENT

A combination of D* and focal exophytic mass can be used to differentiate placenta percreta.

Performance of magnetic resonance imaging to predict maternal outcomes in patients at high risk for placenta accreta spectrum disorder

OBJECTIVE

The purpose of this study is to evaluate the diagnostic performance of MRI parameters to predict adverse maternal peripartum outcomes in pregnant females at high-risk for placenta accreta spectrum (PAS) disorder.

METHODS AND MATERIALS

This retrospective study evaluated 60 pregnant females who underwent MRI for placental assessment. MRI studies were reviewed by a radiologist blinded to all clinical data. MRI parameters were compared with five maternal outcomes: severe bleeding, cesarean hysterectomy, prolonged operation time, need for blood transfusion, and need for intensive care unit (ICU) admission. The MRI findings were associated with pathologic and/or intraoperative findings for PAS.

RESULTS

The study identified 46 cases of PAS disorder and 16 cases of placenta percreta. The agreement between the radiologist impression of PAS disorder and the intraoperative/histological findings was substantial (0.67, p < 0.001), and almost perfect for the presence of placenta percreta (0.87, p < 0.001). The presence of a placental bulge was highly associated with placenta percreta, with sensitivity of 87.5% and specificity of 90.9%. The MRI signs that associated with more maternal outcomes were myometrial thinning, with significant odds ratio for severe blood loss (20.2), hysterectomy (4.0), need for blood transfusion (4.8) and prolonged surgery time (4.9), and uterine bulging, with significant odds ratio for severe blood loss (11.9), hysterectomy (34.0), ICU admission (5.0), and need for blood transfusion (4.8).

CONCLUSION

MRI signs significantly correlated with invasive placenta and were independently associated with adverse maternal outcomes. The presence of a placental bulge was highly accurate in predicting placenta percreta.

ADVANCES IN KNOWLEDGE

First study to evaluate the strength of the association between individual MRI signs and five adverse maternal outcomes. Conclusions support published MRI signs associated with placental invasion, especially regarding the value placental bulging in predicting placenta percreta.

Placenta Accreta Spectrum Disorders and Radiomics: Systematic review and quality appraisal

PURPOSE

Ultrasound and magnetic resonance imaging are the imaging modalities of choice for placenta accrete spectrum (PAS) disorders assessment. Radiomics could further increase the value of medical images and allow to overcome the limitations linked to their visual assessment. Aim of this systematic review was to identify and appraise the methodological quality of radiomics studies focused PAS disorders applications.

METHOD

Three online databases (PubMed, Scopus and Web of Science) were searched to identify original research articles on human subjects published in English. For the qualitative synthesis of results, data regarding study design (e.g., retrospective or prospective), purpose, patient population (e.g., sample size), imaging modalities and radiomics pipelines (e.g., segmentation and feature extraction strategy) were collected. The appraisal of methodological quality was performed using the Radiomics Quality Score (RQS).

RESULTS

10 articles were finally included and analyzed. All were retrospective and MRI-powered. The majority included more than 100 patients (6/10). Four were prognostic (focused on either the prediction of bleeding volume or the prediction of needed management) while six diagnostic (PAS vs not PAS classification) studies. The median RQS was 8, with maximum and minimum respectively equal to 17/36 and - 6/36. Major methodological concerns were the lack of feature stability to multiple segmentation testing and poor data openness.

CONCLUSIONS

Radiomics studies focused on PAS disorders showed a heterogeneous methodological quality, overall lower than desirable. Furthermore, many relevant research questions remain unexplored. More robust investigations are needed to foster advancements in the field and possibly clinical translation.

Maternal Smoking and the Risk of Placenta Accreta Spectrum: A Systematic Review and Meta-Analysis

Background

This is the first meta-analysis that assessed the association between maternal smoking and the risk of placenta accreta spectrum (PAS), so this study was aimed at investigating the association between maternal smoking and PAS based on observational studies. PAS is defined as a severe obstetric complication due to the abnormal invasion of the chorionic villi into the myometrium and uterine serosa.

Methods

We searched electronic bibliographic databases including PubMed, Web of Science, Scopus, Science Direct, and Google Scholar until January 2022. The results were reported using a random effect model. The chi-square test and the I² statistic were used to assess heterogeneity. Egger's and Begg's tests were used to examine the probability of publication bias. All statistical analyses were performed at a significance level of 0.05 using Stata software, version 11.

Results

Based on the random effect model, the estimated OR of the risk of PAS associated with smoking was 1.21 (95% CI: 1.02, 1.41; I² = 4.7%). Subgroup analysis was conducted based on study design, and the result showed that the association between smoking and PAS among cohort studies was significant 1.35 (95% CI: 1.15, 1.55; I² = 0.0%).

Conclusion

Our results suggested that maternal smoking is a risk factor for the PAS. There was no heterogeneity among studies that reported an association between smoking and the PAS. The Newcastle-Ottawa Scale (NOS) was used to measure study quality.

A computerized diagnostic model for automatically evaluating placenta accrete spectrum disorders based on the combined MR radiomics-clinical signatures

We aimed to establish a computerized diagnostic model to predict placenta accrete spectrum (PAS) disorders based on T2-weighted MR imaging. We recruited pregnant women with clinically suspected PAS disorders between January 2015 and December 2018 in our institution. All preoperative T2-weighted imaging (T2WI) MR images were manually outlined on the picture archive communication system terminal server. A nnU-Net network for automatic segmentation and the corresponding radiomics features extracted from the segmented region were applied to build a radiomics-clinical model for PAS disorders identification. Taking the surgical or pathological findings as the reference standard, we compared this computerized model’s diagnostic performance in detecting PAS disorders. In the training cohort, our model combining both radiomics and clinical characteristics yielded an accuracy of 0.771, a sensitivity of 0.854, and a specificity of 0.750 in identifying PAS disorders. In the testing cohort, this model achieved a segmentation mean Dice coefficient of 0.890 and yielded an accuracy of 0.825, a sensitivity of 0.830 and a specificity of 0.822. In the external validation cohort, this computer-aided diagnostic model yielded an accuracy of 0.690, a sensitivity of 0.929 and a specificity of 0.467 in identifying placenta increta. In the present study, a machine learning model based on preoperative T2WI-based imaging had high accuracy in identifying PAS disorders in respect of surgical and histological findings.

Magnetic resonance imaging and previous cesarean section in placenta accrete spectrum disorder: Predictor model

OBJECTIVE

To evaluate objective criteria of Magnetic Resonance Imaging (MRI) of Placenta Accreta Spectrum disorder (PAS) analyzing interobserver agreement and to derive a model including imaging and clinical variables to predict PAS.

METHODS

A retrospective review including patients submitted to MRI with suspicious findings of PAS on ultrasound. Exclusion criteria were lack of pathology or surgical information and missing or poor-quality MRI. Two radiologists analyzed six MRI features, and significant clinical data were also recorded. PAS confirmed on pathology or during intraoperative findings were considered positive for the primary outcome. Variables were tested through logistic regression models.

RESULTS

Final study included 96 patients with a mean age of 33 years and 73.0% of previous C-sections. All MRI features were significantly associated with PAS for both readers. After logistic regression fit, including MRI signs with a moderate or higher interobserver agreement, intraplacental T2 dark band was the most significant radiologic criteria, and ROC analysis resulted in an AUC = 0.782. After including the most relevant clinical data (previous C-section) to the model, the ROC analysis improved to an AUC = 0.893.

CONCLUSION

Simplified objective criteria on MRI, including intraplacental T2 dark band associated with clinical information of previous C-sections, had the highest accuracy and was used for a predictive model of PAS.

Qualitative evaluation of MR images for assessing placenta accreta spectrum disorders in patients with placenta previa: A pilot validation study

PURPOSE

To validate a qualitative imaging method using magnetic resonance (MR) for predicting placental accreta spectrum (PAS) in patients with placenta previa (PP).

METHOD

Two MR imaging methods built in our previous experience was tested in an external comparable group of sixty-five patients with PP; these methods consisted of presence of at least one (Method 1) or two (Method 2) of the following abnormal MR imaging signs: intraplacental dark bands, focal interruption of myometrial border and abnormal placental vascularity. Three groups of radiologists with different level of expertise evaluated MR images: at least 5 years of experience in body imaging (Group 1); at least 10 (Group 2) or 20 (Group 3) years of experience in genito-urinary MR. While radiologists of Group 1 routinely evaluated MR images, those of Groups 2 and 3 used both Methods 1 and 2.

RESULTS

A significant (p < 0.005) difference was found between the diagnostic accuracy values of imaging evaluation performed by Group 3 using Method 1 (63%) and Method 2 (89%); of note, the accuracy of Method 2 by Group 3 was also significantly (p < 0.005) higher compared to that of both Methods 1 (46%) and 2 (63%) by Group 2 as well as to that of the routine evaluation by Group 1 (60%).

CONCLUSIONS

The qualitative identification of at least two abnormal MR signs (Method 2) represents an accurate method for predicting PAS in patients with PP particularly when this method was used by more experienced radiologists; thus, imaging expertise and methodology is required for this purpose.