RESUMO
OBJECTIVE: We previously demonstrated the potential of radiomics for prediction of severe histological Placenta Accreta Spectrum (PAS) subtypes using T2-weighted MRI. We aim is to validate our model using an additional dataset. Secondly, we explore whether performance is improved using a new approach to develop a new multivariate radiomics model. METHODS: Multi-centre retrospective analysis conducted between 2018-2023. Inclusion criteria: MRI performed for suspicion of PAS from ultrasound, clinical findings of PAS at laparotomy and/or histopathological confirmation. Radiomic features were extracted from T2-weighted MRI. The previous multivariate model was validated. Secondly, a 5-radiomic feature random forest classifier was selected from a randomised feature selection scheme to predict invasive placenta increta PAS cases. Prediction performance was assessed based on several metrics including Area Under the Curve (AUC) of the receiver operating characteristic curve (ROC), sensitivity and specificity. RESULTS: We present 100 women (mean age 34.6 (±3.9) with PAS, 64 of whom had placenta increta. Firstly, we validated the previous multivariate model and found a Support Vector Machine classifier had a sensitivity of 0.620 (95% CI: 0.068; 1.0), specificity of 0.619 (95% CI: 0.059; 1.0), an AUC of 0.671 (95% CI: 0.440; 0.922) and accuracy of 0.602 (95% CI: 0.353; 0.817) for predicting placenta increta. From the new multivariate model, the best 5-feature subset selected via the random subset feature selection scheme comprised of 4 radiomic features and 1 clinical variable (number of previous caesareans). This clinical-radiomic model achieved an AUC of 0.713 (95% CI: 0.551; 0.854), accuracy of 0.695 (95% CI 0.563; 0.793), sensitivity of 0.843 (95% CI 0.682; 0.990) and specificity of 0.447 (95% CI 0.167; 0.667). CONCLUSION: We validated our previous model and present a new multivariate radiomic model for prediction of severe placenta increta from a well-defined, cohort of PAS cases. ADVANCES IN KNOWLEDGE: Radiomic features demonstrate good predictive potential for identifying placenta increta. This suggests radiomics may be a useful adjunct to clinicians caring for women with this high-risk pregnancy condition.
RESUMO
Trauma is the leading non-obstetric cause of maternal and fetal mortality and affects an estimated 5-7% of all pregnancies. Pregnant women, thankfully, are a small subset of patients presenting in the trauma bay, but they do have distinctive physiologic and anatomic changes. These increase the risk of certain traumatic injuries, and the gravid uterus can both be the primary site of injury and mask other injuries. The primary focus of the initial management of the pregnant trauma patient should be that of maternal stabilization and treatment since it directly affects the fetal outcome. Diagnostic imaging plays a pivotal role in initial traumatic injury assessment and should not deviate from normal routine in the pregnant patient. Radiographs and focused assessment with sonography in the trauma bay will direct the use of contrast-enhanced computed tomography (CT), which remains the cornerstone to evaluate the potential presence of further management-altering injuries. A thorough understanding of its risks and benefits is paramount, especially in the pregnant patient. However, like any other trauma patient, if evaluation for injury with CT is indicated, it should not be denied to a pregnant trauma patient due to fear of radiation exposure.
RESUMO
BACKGROUND: Placenta accreta spectrum (PAS) is a rare, life-threatening complication of pregnancy. Predicting PAS severity is critical to individualise care planning for the birth. We aim to explore whether radiomic analysis of T2-weighted magnetic resonance imaging (MRI) can predict severe cases by distinguishing between histopathological subtypes antenatally. METHODS: This was a bi-centre retrospective analysis of a prospective cohort study conducted between 2018 and 2022. Women who underwent MRI during pregnancy and had histological confirmation of PAS were included. Radiomic features were extracted from T2-weighted images. Univariate regression and multivariate analyses were performed to build predictive models to differentiate between non-invasive (International Federation of Gynecology and Obstetrics [FIGO] grade 1 or 2) and invasive (FIGO grade 3) PAS using R software. Prediction performance was assessed based on several metrics including sensitivity, specificity, accuracy and area under the curve (AUC) at receiver operating characteristic analysis. RESULTS: Forty-one women met the inclusion criteria. At univariate analysis, 0.64 sensitivity (95% confidence interval [CI] 0.0-1.00), specificity 0.93 (0.38-1.0), 0.58 accuracy (0.37-0.78) and 0.77 AUC (0.56-.097) was achieved for predicting severe FIGO grade 3 PAS. Using a multivariate approach, a support vector machine model yielded 0.30 sensitivity (95% CI 0.18-1.0]), 0.74 specificity (0.38-1.00), 0.58 accuracy (0.40-0.82), and 0.53 AUC (0.40-0.85). CONCLUSION: Our results demonstrate a predictive potential of this machine learning pipeline for classifying severe PAS cases. RELEVANCE STATEMENT: This study demonstrates the potential use of radiomics from MR images to identify severe cases of placenta accreta spectrum antenatally. KEY POINTS: ⢠Identifying severe cases of placenta accreta spectrum from imaging is challenging. ⢠We present a methodological approach for radiomics-based prediction of placenta accreta. ⢠We report certain radiomic features are able to predict severe PAS subtypes. ⢠Identifying severe PAS subtypes ensures safe and individualised care planning for birth.
Assuntos
Placenta Acreta , Gravidez , Humanos , Feminino , Placenta Acreta/diagnóstico por imagem , Estudos Prospectivos , Estudos Retrospectivos , Aprendizado de Máquina , Projetos de PesquisaRESUMO
BACKGROUND: Myocardial fibrosis leads to diastolic dysfunction in patients with hypertrophic cardiomyopathy (HCM). OBJECTIVES: To evaluate a manual method of measuring mitral annular relaxation velocity (termed cardiac MRI e') as a measure of diastolic dysfunction on routine cardiac MRI and its relationship with myocardial late-gadolinium enhancement (LGE) and feature tracking measures of diastolic dysfunction in patients with HCM. METHODS: CMR e', feature tracking measures of diastolic function, left atrial, left ventricular (LV) parameters and LGE were retrospectively measured in 75 patients with HCM (mean age, 54.7 years ± 15.3, 54 men). Multivariate regression and partial Spearman correlations were performed. RESULTS: Cardiac MRI e' measures correlated with LGE (r = 0.49, P < 0.001) and multiple feature tracking measures of diastolic function, adjusted for patient demographics, left atrial and left ventricular parameters. Cardiac MRI e' measures were independently predictive of LGE ≥ 10% (mean total cardiac MRI e': LGE < 10% vs LGE ≥ 10% was 3.5 cm/s vs. 1.7 cm/s, P < 0.001). Superior CMR e' had an AUC of 0.79 [95%CI 0.66-0.92, P < 0.0001]) in predicting patients with LGE ≥ 10% and a cutoff of 1.7 cm/s resulted in a sensitivity and specificity of 81.0% and 78.0% respectively. CONCLUSION: Cardiac MRI e' is a manual measure of LV diastolic dysfunction acquired on routine cardiac MRI without specialized software and is an independent predictor of LGE ≥ 10% and diastolic dysfunction in HCM.
Assuntos
Cardiomiopatia Hipertrófica , Gadolínio , Cardiomiopatia Hipertrófica/diagnóstico por imagem , Meios de Contraste , Fibrose , Humanos , Imageamento por Ressonância Magnética/métodos , Masculino , Pessoa de Meia-Idade , Estudos RetrospectivosRESUMO
BACKGROUND: Differentiating combined pulmonary fibrosis with emphysema (CPFE) from pure emphysema can be challenging on high-resolution computed tomography (HRCT). This has antifibrotic therapy implications. METHODS: Twenty patients with suspected CPFE underwent late gadolinium-enhanced (LGE) thoracic magnetic resonance imaging (LGE-MRI) and HRCT. Data from twelve healthy control subjects from a previous study who underwent thoracic LGE-MRI were included for comparison. Quantitative LGE signal intensity (SI) was retrospectively compared in regions of fibrosis and emphysema in CPFE patients to similar lung regions in controls. Qualitative comparisons for the presence/extent of reticulation, honeycombing, and traction bronchiectasis between LGE-MRI and HRCT were assessed by two readers in consensus. RESULTS: There were significant quantitative differences in fibrosis SI compared to emphysema SI in CPFE patients (25.8, IQR 18.4-31.0 versus 5.3, IQR 5.0-8.1, p < 0.001). Significant differences were found between LGE-MRI and HRCT in the extent of reticulation (12.5, IQR 5.0-20.0 versus 25.0, IQR 15.0-26.3, p = 0.038) and honeycombing (5.0, IQR 0.0-10.0 versus 20.0, IQR 10.6-20.0, p = 0.001) but not traction bronchiectasis (10.0, IQR 5-15 versus 15.0, IQR 5-15, p = 0.878). Receiver operator curve analysis of fibrosis SI compared to similarly located regions in control subjects showed an area under the curve of 0.82 (p = 0.002). A SI cutoff of 19 yielded a sensitivity of 75% and specificity of 86% in differentiating fibrosis from similarly located regions in control subjects. CONCLUSION: LGE-MRI can differentiate CPFE from pure emphysema and may be a useful adjunct test to HRCT in patients with suspected CPFE.