Few-shot learning to identify atypical endometrial hyperplasia and endometrial cancer based on transvaginal ultrasonic images.

Objective: It is challenging to accurately distinguish atypical endometrial hyperplasia (AEH) and endometrial cancer (EC) under routine transvaginal ultrasonic (TVU) detection. Our research aims to use the few-shot learning (FSL) method to identify non-atypical endometrial hyperplasia (NAEH), AEH, and EC based on limited TVU images. Methods: The TVU images of pathologically confirmed NAEH, AEH, and EC patients (n = 33 per class) were split into the support set (SS, n = 3 per class) and the query set (QS, n = 30 per class). Next, we used dual pretrained ResNet50 V2 which pretrained on ImageNet first and then on extra collected TVU images to extract 1*64 eigenvectors from the TVU images in SS and QS. Then, the Euclidean distances were calculated between each TVU image in QS and nine TVU images of SS. Finally, the k-nearest neighbor (KNN) algorithm was used to diagnose the TVU images in QS. Results: The overall accuracy and macro precision of the proposed FSL model in QS were 0.878 and 0.882 respectively, superior to the automated machine learning models, traditional ResNet50 V2 model, junior sonographer, and senior sonographer. When identifying EC, the proposed FSL model achieved the highest precision of 0.964, the highest recall of 0.900, and the highest F1-score of 0.931. Conclusions: The proposed FSL model combining dual pretrained ResNet50 V2 eigenvectors extractor and KNN classifier presented well in identifying NAEH, AEH, and EC patients with limited TVU images, showing potential in the application of computer-aided disease diagnosis.

Evaluation of fetal cardiac morphology and function by fetal heart quantification technique in the normal second and third trimesters.

Background: The study of fetal heart is receiving increasing attention. Fetal heart quantification (Fetal HQ) technology is a new speckle tracking technology that can analyze the 24-segment morphology and function of fetal ventricles. This study aims to use Fetal HQ to assess the changes in the structure and function of the fetal heart in normal mid to late pregnancy, providing a foundation for the clinical application of fetal cardiac speckle tracking technology. Methods: The heart size, global sphericity index (GSI), left ventricular [stroke volume (SV)], cardiac output (CO), ejection fraction (EF), global longitudinal strain (GLS), fractional area change (FAC), 24-segment end-diastolic diameter (EDD), sphericity index (SI) and fractional shortening (FS) of the two ventricles of 500 normal pregnant fetuses were evaluated by Fetal HQ. The subjects were divided into 5 groups according to gestational weeks (GA), and the changes of fetal heart morphology and function were observed. P<0.05 indicated the statistically significant difference. Results: The fetal heart rate decreased gradually with the increase of GA (P<0.05). The size parameters of the fetal heart and two ventricles gradually increased with increasing GA (P<0.05). The 24 segments EDD of both ventricles increased with increasing GA (P<0.05), while the EDD increased first and then decreased from the ventricle base to the apex. The GSI and the 24 segments SI of two ventricles were basically not significantly different among the groups (P>0.05). The EF, GLS, FAC of the left ventricle and the GLS, FAC of the right ventricle decreased with the increase of GA (P<0.05), and SV and CO increased with increasing GA (P<0.05). The 24 segments FS of the left ventricle showed a downward trend with the increase of GA and gradually increased from the base to the apex. The FS of most segments of the right ventricle decreased with the increase of the GA and increased first and then decreased from the base to the apex. Conclusions: The whole and segmental size parameters of fetal heart can quantitatively evaluate the growth and development of fetal heart; the GSI and segmental SI are reliable morphological indexes for evaluating fetal heart; fetal ventricular function parameters EF, FAC, GLS and segmental FS can evaluate fetal cardiac function. The Fetal HQ technique can help us to evaluate the heart growth and development of normal fetuses in the second and third trimester of pregnancy.

Relationship Between Subclavian Artery Stenosis Lesions and Posterior Circulation Infarction: A Preliminary Study.

BACKGROUND: To investigate the correlation between subclavian steal syndrome and posterior circulation infarction using magnetic resonance imaging. METHODS: A total of 294 patients diagnosed with subclavian steal syndrome using carotid Doppler ultrasonography were retrospectively included. According to the magnetic resonance imaging results, they were divided into posterior circulation infarction group and nonposterior circulation infarction group. Clinical indicators and carotid Doppler ultrasound parameters of patients were collected, and they were screened to establish a multiple logistic regression model. Receiver operating characteristic curve analysis of the established multiple logistic regression model was performed, and the area under the curve was calculated to evaluate the predictive efficiency of the model. RESULTS: After statistical analysis of all parameters of the 2 groups of patients, a total of 10 parameters were included in multiple logistic regression to establish a model. The results showed a correlation between posterior circulation infarction and subclavian artery occlusion, grade III subclavian steal syndrome, gender, vulnerable plaques, National Institutes of Health Stroke Scale score, and age. After the receiver operating characteristic curve analysis of the model, the area under the curve for the multiple logistic regression model was 0.773. CONCLUSIONS: The multiparameter composite model based on clinical baseline data and carotid Doppler ultrasonography parameters can effectively predict posterior circulation infarction and offer novel insight for clinical diagnosis.

A multiplex RPA-CRISPR/Cas12a-based POCT technique and its application in human papillomavirus (HPV) typing assay.

Persistent infection with high-risk human papillomavirus (HR-HPV) is the primary and initiating factor for cervical cancer. With over 200 identified HPV types, including 14 high-risk types that integrate into the host cervical epithelial cell DNA, early determination of HPV infection type is crucial for effective risk stratification and management. Presently, on-site immediate testing during the HPV screening stage, known as Point of Care Testing (POCT), remains immature, severely limiting the scope and scenarios of HPV screening. This study, guided by the genomic sequence patterns of HPV, established a multiplex recombinase polymerase amplification (RPA) technology based on the concept of "universal primers." This approach achieved the multiple amplification of RPA, coupled with the CRISPR/Cas12a system serving as a medium for signal amplification and conversion. The study successfully constructed a POCT combined detection system, denoted as H-MRC12a (HPV-Multiple RPA-CRISPR/Cas12a), and applied it to high-risk HPV typing detection. The system accomplished the typing detection of six high-risk HPV types (16, 18, 31, 33, 35, and 45) can be completed within 40 min, and the entire process, from sample loading to result interpretation, can be accomplished within 45 min, with a detection depth reaching 1 copy/µL for each high-risk type. Validation of the H-MRC12a detection system's reproducibility and specificity was further conducted through QPCR on 34 clinical samples. Additionally, this study explored and optimized the multiplex RPA amplification system and CRISPR system at the molecular mechanism level. Furthermore, the primer design strategy developed in this study offers the potential to enhance the throughput of H-MRC12a detection while ensuring sensitivity, providing a novel research avenue for high-throughput detection in Point-of-Care molecular pathogen studies.

Ultrasonographic diagnosis of ovarian tumors through the deep convolutional neural network.

OBJECTIVES: The objective of this study was to develop and validate an ovarian tumor ultrasonographic diagnostic model based on deep convolutional neural networks (DCNN) and compare its diagnostic performance with that of human experts. MATERIAL AND METHODS: We collected 486 ultrasound images of 192 women with malignant ovarian tumors and 617 ultrasound images of 213 women with benign ovarian tumors, all confirmed by pathological examination. The image dataset was split into a training set and a validation set according to a 7:3 ratio. We selected 5 DCNNs to develop our model: MobileNet, Xception, Inception, ResNet and DenseNet. We compared the performance of the five models through the area under the curve (AUC), sensitivity, specificity, and accuracy. We then randomly selected 200 images from the validation set as the test set. We asked three expert radiologists to diagnose the images to compare the performance of radiologists and the DCNN model. RESULTS: In the validation set, AUC of DenseNet was 0.997 while AUC was 0.988 of ResNet, 0.987 of Inception, 0.968 of Xception and 0.836 of MobileNet. In the test set, the accuracy was 0.975 with the DenseNet model versus 0.825 (p < 0.0001) with the radiologists, and sensitivity was 0.975 versus 0.700 (p < 0.0001), and specificity was 0.975 versus 0.908 (p < 0.001). CONCLUSIONS: DensNet performed better than other DCNNs and expert radiologists in identifying malignant ovarian tumors from benign ovarian tumors based on ultrasound images, a finding that needs to be further explored in clinical trials.

Retrospective validation of ultrasound characteristics at 19-24 weeks as predictive tools for selective intrauterine growth restriction in monochorionic diamniotic twin pregnancies: a diagnostic study.

Background: Seeking an optimal time point for ultrasound examination is important for the diagnosis of late selective intrauterine growth restriction (sIUGR) at birth in monochorionic diamniotic (MCDA) twin pregnancies. We aimed to assess the role of ultrasound characteristics at 19-24 weeks as predictive tools for late sIUGR at birth in MCDA twin pregnancies. Methods: We retrospectively recruited 32 sIUGR and 56 normal patients with MCDA twin pregnancies. Ultrasound indexes of these included subjects at 19-24 weeks, including the middle cerebral artery peak systolic velocity (MCA-PSV), umbilical artery pulsatility index (UA-PI), middle cerebral artery pulsatility index (MCA-PI), and cerebroplacental ratio (CPR) were assessed. Receiver operating characteristic (ROC) curves were used to ascertain the predictive value of ultrasound characteristics discrepancy for such complications, and the relationship between the ultrasound characteristics and sIUGR was assessed by a logistic regression analysis. Results: Differences were found in the MCA-PI, UA-PI, and CPR discordances between the normal MCDA and sIUGR subjects. CPR discordance was the most effective characteristic for predicting sIUGR [area under the ROC curve (AUC) =0.883; 95% CI: 0.795-0.948], followed by UA-PI discordance (AUC =0.772; 95% CI: 0.685-0.829), and MCA-PI discordance (AUC =0.746; 95% CI: 0.681-0.823), respectively. Additionally, the optimal cutoff value of CPR discordance was 21.65, and the corresponding sensitivity and specificity were 0.750 and 0.929, respectively. The correlation analysis revealed that gestational age (GA) at ultrasound scan but not at delivery was significantly correlated with the MCA-PSV (r=0.55, P<0.01), UA-PI (r=0.55, P<0.01), MCA-PI (r=0.49, P<0.01), and CPR (r=0.55, P<0.01) in sIUGR, while GA at both ultrasound scan and birth was significantly correlated with MCA-PSV (r=0.65, P<0.01), UA-PI (r=0.49, P<0.01), MCA-PI (r=0.48, P<0.01), and CPR (r=0.63, P<0.01) in normal MCDA. Conclusions: Increased MCA-PI, UA-PI, and CPR discordances were found in fetuses with sIUGR. CPR discordance could serve as a predictive index for sIUGR. An early ultrasound examination may be more accurate than biochemical modality for sIUGR prediction.

Assessment of myocardial injury in neonates born to pregnant women with pregnancy complicated by severe preeclampsia by myocardial work indices: a prospective study. / å¿ƒè‚ŒåšåŠŸå‚æ•°è¯„ä¼°é‡åº¦å­ç—«å‰æœŸå­•å¦‡å­ä»£æ–°ç”Ÿå„¿æœŸå¿ƒè‚ŒæŸä¼¤çš„å‰çž»æ€§ç ”ç©¶.

OBJECTIVES: To evaluate myocardial injury in neonates born to pregnant women with pregnancy complicated by severe preeclampsia by myocardial work indices. METHODS: A prospective cohort study was performed on 25 preterm infants born to the pregnant women with severe preeclampsia from June 2020 to April 2021 (severe preeclampsia group), and 25 preterm infants born to the pregnant women without severe complications in pregnancy were enrolled as the control group. Echocardiography was performed within 24 hours and at 48-72 hours and 14-28 days after birth to measure conventional parameters. Two-dimensional speckle-tracking echocardiography was performed to construct a noninvasive left ventricular pressure-strain loop based on two-dimensional myocardial strain and left ventricular systolic pressure noninvasively measured, so as to calculate myocardial work indices. RESULTS: Compared with the control group, the severe preeclampsia group had significant reductions in left ventricular global work index and global constructive work within 24 hours after birth (P<0.05), a significant reduction in left ventricular global work efficiency and a significant increase in global waste work at 48-72 hours after birth (P<0.05), and a significant reduction in left ventricular global work efficiency at 14-28 days after birth (P<0.05). CONCLUSIONS: Subclinical myocardial injury persists in the neonatal period in preterm infants born to pregnant women with severe preeclampsia.

Retrospective validation of 11-13 weeks' gestation ultrasound characteristics as predictive tools for twin-twin transfusion syndrome and selective intrauterine growth restriction in monochorionic diamniotic twin pregnancies.

BACKGROUND: Twin to twin transfusion syndrome (TTTS) is a serious syndrome that can affect twin pregnancies involving a single placenta, impacts some of twin gestations with monochorionic diamniotic (MCDA) placentas. We validated the ultrasound characteristics of 11-13 weeks' gestation to predict TTTS and selective intrauterine growth restriction (sIUGR) in MCDA pregnancies. METHODS: We retrospectively included all of the MCDA twin pregnancies with ultrasound characteristics, including the crown-rump length (CRL), ductus venosus pulsatility index for veins (DV PIV), and nuchal translucency (NT) thickness, at 11-13 weeks' gestation, followed by mean difference and discordance comparison. Receiver operating characteristic (ROC) curves were constructed for the comparison of values of these predictive markers for identification of MCDA pregnancies with high-risk of adverse outcomes. RESULTS: A total of 98 MCDA pregnancies were included in this study. Among the 98, 34 (34.7%) developed sIUGR, whereas 10 (10.2%) expressed TTTS. Significant differences in NT discordance were found among the normal, sIUGR, and TTTS groups; moreover, a significant difference was found between pregnancies with normal outcomes and sIUGR (P<0.001), normal and TTTS (P<0.001), and sIUGR and TTTS (P<0.001). Difference in NT was determined to be the best predictive marker for sIUGR [area under the curve (AUC) =0.769; 95% confidence interval (CI): 0.591 to 0.992], and NT discordance was considered the best predictive marker for TTTS (AUC =0.802; 95% CI: 0.485 to 0.936). CONCLUSIONS: Significant differences in NT discordance were found between the normal, sIUGR, and TTTS groups, while NT difference and NT discordance were identified as predictive markers for sIUGR and TTTS, respectively.

UBE2S promotes the progression and Olaparib resistance of ovarian cancer through Wnt/ß-catenin signaling pathway.

BACKGROUND: Ovarian cancer is the most lethal gynecologic malignancy worldwide. Olaparib, an inhibitor of poly (ADP-ribose) polymerase (PARP), is becoming widely used in ovarian cancer treatment. The overall survival of ovarian cancer has not been significantly changed over the past decades and ovarian cancer has become increasingly resistant to the Olaparib. Ubiquitin-conjugating enzyme E2S (UBE2S) has been proved to promote malignant behaviors in many cancers. However, the function of UBE2S in the development and Olaparib resistance of ovarian cancer are unclear. MATERIALS AND METHODS: In this study, we detected the expression of UBE2S in normal fallopian tube (FT) and HGSOC tissues. A2780 and SKOV3 cells were stably transfected with PCMV-UBE2S, PCMV-UBE2S-C95S, UBE2S shRNAs, and negative controls. The CCK8 assay and clonogenic assay were conducted to analyze ovarian cancer proliferation and Olaparib resistance. The transwell assay was performed to determine the migration and invasion of ovarian cancer cells. The relative protein levels of the Wnt/ß-catenin signaling pathway were tested using western blot. The ovarian cancer cells were treated with XAV-939 to investigate the role of Wnt/ß-catenin signaling pathway in Olaparib resistance. Moreover, we repeated some above procedures in the xenograft model. RESULTS: The results demonstrated that UBE2S was highly upregulated in HGSOC and that high UBE2S expression was correlated with poor outcomes in HGSOC. UBE2S promoted ovarian cancer proliferation and drived the migration and invasion of ovarian cancer cells. UBE2S activated the Wnt/ß-catenin signaling pathway in ovarian cancer resulting in Olaparib resistance in vitro and in vivo. Furthermore, UBE2S enhanced the proliferation and Olaparib resistance of ovarian cancer in its enzymatic activity dependent manner. CONCLUSIONS: These data suggest a possible molecular mechanism of proliferation and metastasis of ovarian cancer and highlight the potential role of UBE2S as a therapeutic target in ovarian cancer.

Myocardial Function in Offspring Aged 5 to 8 Years of Pregnancy Complicated by Severe Preeclampsia Measured by Two-Dimensional Speckle-Tracking Echocardiography.

Objective: This study aimed to quantitatively assess myocardial strain in preterm children aged 5 to 8 years of pregnancy complicated by severe preeclampsia (PE) by two-dimensional (2D) speckle tracking echocardiography. Method: A cohort study of 23 preterm children delivered by severe PE pregnant women from 2010 to 2012 in the First Affiliated Hospital of Soochow University was carried out. 23 preterm children from uneventful pregnancies in the same period served as controls. Myocardial functions including left ventricular longitudinal strain, radial strain, circumferential strain, and right ventricular longitudinal strain were evaluated by conventional Doppler, tissue Doppler imaging, and 2D speckle-tracking echocardiography (2D STE). All examinations were performed by an experienced ultrasonographer using the VIVID E9 (GE Healthcare) machine, according to standard techniques. Results: Children aged 5-8 years delivered from severe PE presented less weight (24.41 vs. 20.89 kg, P < 0.05), shorter height (124.1 vs 115.6 cm, P < 0.05) and faster heart rates (84 vs. 93 bpm, P < 0.05) compared to offspring of normotensive women. There were no significant differences in global left ventricular longitudinal strain, radial strain, circumferential strain, and right ventricular longitudinal strain between the children in the experimental group and the control group (P > 0.05). Conclusion: Exposure to the intrauterine environment of severe PE during the fetal period did not have a significant impact on cardiac structure in premature children at 5-8 years old, but they had a higher resting heart rate which may be associated with cardiovascular disease in the long run.