Dataset Information

Grade Prediction of Bleeding Volume in Cesarean Section of Patients With Pernicious Placenta Previa Based on Deep Learning.

ABSTRACT: In order to predict the amount of bleeding in the cesarean section of the patients with Pernicious Placenta Previa (PPP), this study proposed an automatic blood loss prediction method based on Magnetic Resonance Imaging (MRI) uterus image. Firstly, the DeepLab-V3 + network was used to segment the original MRI abdominal image to obtain the uterine region image. Then, the uterine region image and the corresponding blood loss data were trained by Visual Geometry Group Network-16 (VGGNet-16) network. The classification model of blood loss level was obtained. Using a dataset of 82 positive samples and 128 negative samples, the proposed method achieved accuracy, sensitivity and specificity of 75.61, 73.75, and 77.46% respectively. The experimental results showed that this method can not only automatically identify the uterine region of pregnant women, but also objectively determine the level of intraoperative bleeding. Therefore, this method has the potential to reduce the workload of the attending physician and improve the accuracy of experts' judgment on the level of bleeding during cesarean section, so as to select the corresponding hemostasis measures.

SUBMITTER: Liu J

PROVIDER: S-EPMC7203465 | biostudies-literature | 2020

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Grade Prediction of Bleeding Volume in Cesarean Section of Patients With Pernicious Placenta Previa Based on Deep Learning.

Liu Jun J Wu Tao T Peng Yun Y Luo Rongguang R

Frontiers in bioengineering and biotechnology 20200430

In order to predict the amount of bleeding in the cesarean section of the patients with Pernicious Placenta Previa (PPP), this study proposed an automatic blood loss prediction method based on Magnetic Resonance Imaging (MRI) uterus image. Firstly, the DeepLab-V3 + network was used to segment the original MRI abdominal image to obtain the uterine region image. Then, the uterine region image and the corresponding blood loss data were trained by Visual Geometry Group Network-16 (VGGNet-16) network ...[more]

PMID: 32426340

Similar Datasets

Project description:Background: Pernicious placenta previa (PPP) is one of the most dangerous complications in pregnancy after cesarean section, with high perinatal mortality. This study aimed to develop a nomogram to predict postpartum hemorrhage in patients with PPP. Methods: A total of 246 patients with confirmed PPP at Shengjing Hospital of China Medical University from January 2018 to December 2021 were included. Patients were divided into to two cohorts depending on a postpartum blood loss of > 1000 ml (n = 146) or ≤ 1000 ml (n = 100). Lasso regression analysis was performed on the risk factors screened by univariate analysis to screen out the final risk factors affecting postpartum hemorrhage. Based on the final risk factors, a Nomogram prediction model with excellent performance was constructed using Logistic regression. A nomogram was constructed with further screening of the selected risk factors of postpartum hemorrhage in PPP. A second nomogram based only on the total ultrasonic risk score was constructed. Decision curve analysis (DCA) was used to evaluate the clinical efficacy of the nomograms. Results: Older age, larger gestational age, larger neonatal birth weight, presence of gestational diabetes mellitus, larger amniotic fluid index, absence of gestational bleeding, and higher ultrasonic risk single score were selected to establish a nomogram for postpartum hemorrhage in PPP. The area under the curve of the nomogram constructed by Lasso regression analysis was higher than that of the ultrasonic total score alone (0.887 vs. 0.833). Additionally, DCA indicated better clinical efficacy in the former nomogram than in the later nomogram. Furthermore, internal verification of the nomogram constructed by Lasso regression analysis showed good agreement between predicted and actual values. Conclusion: A nomogram for postpartum hemorrhage in PPP was developed and validated to assist clinicians in evaluating postpartum hemorrhage. This nomogram was more accurate than using the ultrasonic score alone.

Project description:BackgroundPlacenta previa, one of the most severe obstetric complications, carries an increased risk of intraoperative massive hemorrhage. Several risk factors for intraoperative hemorrhage have been identified to date. However, the correlation between birth weight and intraoperative hemorrhage has not been investigated. Here we estimate the correlation between birth weight and the occurrence of intraoperative massive hemorrhage in placenta previa.Materials and methodsWe included all 256 singleton pregnancies delivered via cesarean section at our hospital because of placenta previa between 2003 and 2015. We calculated not only measured birth weights but also standard deviation values according to the Japanese standard growth curve to adjust for differences in gestational age. We assessed the correlation between birth weight and the occurrence of intraoperative massive hemorrhage (>1500 mL blood loss). Receiver operating characteristic curves were constructed to determine the cutoff value of intraoperative massive hemorrhage.ResultsOf 256 pregnant women with placenta previa, 96 (38%) developed intraoperative massive hemorrhage. Receiver-operating characteristic curves revealed that the area under the curve of the combination variables between the standard deviation of birth weight and intraoperative massive hemorrhage was 0.71. The cutoff value with a sensitivity of 81.3% and specificity of 55.6% was -0.33 standard deviation. The multivariate analysis revealed that a standard deviation of >-0.33 (odds ratio, 5.88; 95% confidence interval, 3.04-12.00), need for hemostatic procedures (odds ratio, 3.31; 95% confidence interval, 1.79-6.25), and placental adhesion (odds ratio, 12.68; 95% confidence interval, 2.85-92.13) were independent risk of intraoperative massive hemorrhage.ConclusionIn patients with placenta previa, a birth weight >-0.33 standard deviation was a significant risk indicator of massive hemorrhage during cesarean section. Based on this result, further studies are required to investigate whether fetal weight estimated by ultrasonography can predict hemorrhage during cesarean section in patients with placental previa.

Project description:Objective: To evaluate the efficacy and safety of parallel loop binding compression suture of the lower uterus during cesarean section in pernicious placenta previa complicated with placenta increta. Methods: This retrospective study was performed in patients with pernicious placenta previa complicated with placenta increta or percreta between November 2014 and December 2020 at the Qilu Hospital of Shandong University. Patients underwent parallel loop binding compression suture surgery were defined as study group, and patients underwent traditional surgery with figure-of-eight sutures as the main hemostatic method were defined as control group. Postpartum hemorrhage was evaluated as the primary outcome. The secondary outcomes included age, gestational weeks, operative time, fetal childbirth time, prevention of hysterectomy, blood transfusion, duration of postoperative catheterization, duration of antibiotic treatment, and postoperative hospitalization (days). Additionally, neonatal outcomes were evaluated. Results: A total of 124 patients were enrolled in the study, including 38 patients receiving parallel loop binding compression suture surgery in the study group, and 86 patients in the control group. With parallel loop binding compression suture, the average operation time was significantly reduced (109.0 ± 33.5 vs. 134.4 ± 54.2 min, p = 0.00), and the volume of blood lost were also decreased (2152.6 ± 1169.4 vs. 2960.5 ± 1963.6 ml, p = 0.02), which correspondingly reduced RBC transfusion (7.2 ± 3.5 vs. 10.3 ± 8.7 units, p = 0.03) and FFP transfusion (552.6 ± 350.3 vs. 968.0 ± 799.8 ml, p = 0.00). The fetal childbirth time was extended (14.1 ± 5.6 vs. 11.0 ± 8.0 min, p = 0.03), however, there was no increase in NICU admission rates (36.9 vs. 34.9%, p = 0.83). Except for one premature infant (32 weeks) death in the control group, all infants at our hospital were safely discharged after treatment. Conclusion: Parallel loop binding compression suture is an effective, swift, practical, and safe method to reduce postpartum bleeding in women with pernicious placenta previa, complicated with placenta increta. Besides, it has no adverse effects on newborns.

Project description:BackgroundPlacental accreta spectrum disorders (PAS) are a high-risk group for severe postpartum hemorrhage (SPPH), with the incidence of PAS increasing annually. Given that cesarean section and anterior placenta previa are the primary risk factors for PAS, therefore, our study aims to investigate the predictive value of clinical characteristics and ultrasound indicators for SPPH in patients with anterior placenta previa combined with previous cesarean section, providing a theoretical basis for early prediction of SPPH.MethodsA total of 450 patients with anterior placenta previa combined with previous cesarean section were retrospectively analyzed at Shengjing Hospital affiliated with China Medical University between January 2018 and March 2022. Clinical data and ultrasound indicators were collected. Patients were categorized into SPPH (blood loss >2000mL, 182 cases) and non-SPPH (blood loss ≤ 2000mL, 268 cases) groups based on the blood loss within 24 h postpartum. The population was randomly divided into training and validation cohorts at a 7:3 ratio. LASSO and multifactorial logistic regression analyses were utilized to identify independent risk factors for SPPH. Accordingly, a nomogram prediction model was constructed, the predictive performance was assessed using receiver operating characteristic (ROC) curves, calibration curves and decision curve analysis (DCA).ResultsAmong the 450 patients, 182 experienced SPPH (incidence rate, 40.44%). Preoperative systemic immune-inflammatory index, preoperative D-dimer level, preoperative placenta accreta spectrum ultrasound scoring system (PASUSS) score, and one-step-conservative surgery were identified as independent risk factors for SPPH in patients with anterior placenta previa combined with previous cesarean section. A nomogram was constructed based on these factors. The areas under the ROC curves for the training and validation cohorts were 0.844 (95%CI: 0.801-0.888) and 0.863 (95%CI: 0.803-0.923), respectively. Calibration curves and DCA indicated that this nomogram demonstrated good predictive accuracy.ConclusionsThis nomogram presents an effective and convenient prediction model for identifying SPPH in patients with anterior placenta previa combined with previous cesarean section. It can guide surgical planning and improve prognosis.

Dataset Information

Grade Prediction of Bleeding Volume in Cesarean Section of Patients With Pernicious Placenta Previa Based on Deep Learning.

Publications

Grade Prediction of Bleeding Volume in Cesarean Section of Patients With Pernicious Placenta Previa Based on Deep Learning.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets