Search Results (6)

Background/Objectives: Functional assessment of the fetal heart remains a significant challenge in contemporary perinatology due to the absence of a universally accepted gold standard for such evaluations. The aim of this study was to evaluate the reproducibility of parameters derived from two-dimensional speckle-tracking echocardiography (2D STE) using the FetalHQ software. Methods: We enrolled 87 pregnant women between 19 and 23 weeks of gestation who were undergoing mid-trimester screening at the Provincial Hospital Complex in Kielce. Two independent operators acquired 5 s cine-loops of four-chamber views (4CVs) according to a standardized protocol. Reproducibility was assessed by examining intra- and interobserver variability using the intraclass correlation coefficient (ICC) for several cardiac parameters, including the global sphericity index (GSI), global longitudinal strain (GLS), stroke volume (SV), and fractional area change (FAC). Results: Reproducibility varied substantially across the assessed parameters. The highest intraobserver reproducibility was observed for the 4CV GSI (ICC > 0.9). Moderate intraobserver reproducibility (ICCs ranging from 0.5 to 0.75) was noted for left ventricular (LV) parameters, such as end-diastolic area, end-systolic area, end-diastolic volume, and end-systolic volume. Interobserver variability demonstrated higher ICC values, with excellent reproducibility (ICC > 0.9) for the 4CV GSI and LV volume measurements. Good reproducibility (ICCs between 0.75 and 0.9) was observed for specific left ventricular segmental strain indices, whereas other parameters showed moderate to poor reproducibility (ICC < 0.5). Conclusions: Two-dimensional speckle-tracking echocardiography (2D STE) using FetalHQ exhibits variable reproducibility, which is influenced by the choice of parameters, operator experience, and technical factors. This method holds potential for fetal cardiac assessment; however, additional research is required to improve its predictive accuracy and streamline the evaluation process for routine clinical application. Full article

Background/Objectives: The automatic left ventricle segmentation in transesophageal echocardiography (TEE) is of significant importance. In this paper, we constructed a large-scale TEE apical four-chamber view (A4CV) image dataset and proposed an automatic left ventricular segmentation method for the TEE A4CV based on the UNeXt deep neural network. Methods: UNeXt, a variant of U-Net integrating a multilayer perceptron, was employed for left ventricle segmentation in the TEE A4CV because it could yield promising segmentation performance while reducing both the number of network parameters and computational complexity. We also compared the proposed method with U-Net, TransUNet, and Attention U-Net models. Standard TEE A4CV videos were collected from 60 patients undergoing cardiac surgery, from the onset of anesthesia to the conclusion of the procedure. After preprocessing, a dataset comprising 3000 TEE images and their corresponding labels was generated. The dataset was randomly divided into training, validation, and test sets in an 8:1:1 ratio on the patient level. The training and validation sets were used to train the UNeXt, U-Net, TransUNet, and Attention U-Net models for left ventricular segmentation. The dice similarity coefficient (DSC) and Intersection over Union (IoU) were used to evaluate the segmentation performance of each model, and the Kruskal–Wallis test was employed to analyze the significance of DSC differences. Results: On the test set, the UNeXt model achieved an average DSC of 88.60%, outperforming U-Net (87.76%), TransUNet (85.75%, p < 0.05), and Attention U-Net (79.98%; p < 0.05). Additionally, the UNeXt model had a smaller number of parameters (1.47 million) and floating point operations (2.28 giga) as well as a shorter average inference time per image (141.73 ms), compared to U-Net (185.12 ms), TransUNet (209.08 ms), and Attention U-Net (201.13 ms). The average IoU of UNeXt (77.60%) was also higher than that of U-Net (76.61%), TransUNet (77.35%), and Attention U-Net (68.86%). Conclusions: This study pioneered the construction of a large-scale TEE A4CV dataset and the application of UNeXt to left ventricle segmentation in the TEE A4CV. The proposed method may be used for automatic segmentation of the left ventricle in the TEE A4CV. Full article

Objective. Our objective was to assess the proportion of false-positive CHD cases at the first-trimester evaluation of the fetal heart, performed by experienced operators. Methods. This multicenter retrospective study included of pregnant women with suspicion of CHDs during first-trimester screening for aneuploidies. In all cases, the fetal heart assessments were performed by obstetricians with extensive experience in first-trimester scanning, following an extended protocol proposed by SIEOG national guidelines, which included an axial view of the fetal abdomen and chest to assess visceral situs and evaluation of the four-chamber view (4CV) and three-vessel trachea view (3VTV) with color Doppler. In all suspected cases, fetal echocardiography was offered within 16 and/or at 19–22 weeks’ gestation. Results. From a population of 4300 fetuses, 46 CHDs were suspected. Twenty-four cases were excluded from this analysis because the parents opted for early termination of the pregnancies due to associated structural and/or genetic anomalies. For the remaining 22, echocardiography was performed by 16 weeks in 14 cases (64%) and after 16 weeks in 8 cases. In 19 cases (86.4%), a fetal cardiologist confirmed the presence of a CHD. In three cases (13%), the cardiac anatomy was found to be normal at the fetal echocardiography and postnatally. Conclusions. This study shows that the proportion of false-positive cases at the first-trimester ultrasound examination of the fetal heart, performed by experienced operators, may carry a higher risk of false-positive diagnosis than expected. Therefore, this issue must be discussed in instances where a CHD is suspected at the first-trimester screening. Full article

(1) Background: There is no reliable way to assess antenatal fetal pulmonary hypoplasia; however, the biological parameters of the fetal lung can help in evaluating fetal lung development. This study aimed to establish the reference intervals for normal fetal lung biological parameters at 21–40 weeks among the Chinese population. (2) Methods: This was a cross-sectional study of Chinese groups, and included a total of 1388 normal single pregnant women at 21–40 weeks’ gestation. We selected 2134 images of a standard four-chamber view (4CV). ImageJ software (Release 2.14.0) was used to measure the left and right lung areas using a manual tracing method; the elliptic function key was used to measure the fetal thoracic circumference (TC), thoracic area (TA), head circumference (HC), heart area (HA), and abdominal circumference (AC). Based on the above measurements, the following parameters were calculated: lung area to head circumference ratio (LHR), total lung area (TLA), TLA/Weight (mm²/g), cardiothoracic ratio (CTR), lung–thoracic area ratio (TLA/TA), lung–heart area ratio (TLA/HA), TC/AC, and TC/HC. (3) Results: The left and right lung areas and LHRs positively correlated with gestational age (R² = 0.85, 0.88, 0.66, 0.71, p < 0.001). From 21–40 weeks, the left and right lung areas and TLA increased by about 3.33 times, 3.16 times, and 3.22 times, respectively. The means of left and right LHRs increased by about 1.94 times and 1.84 times, respectively. TLA/Weight (mm²/g) was weakly correlated with gestational age, while CTR, TLA/TA, TLA/HA, TC/AC, and TC/HC had no significant correlation with gestational age. There was no statistically significant difference in fetal lung parameters between different genders of newborns, p > 0.05. (4) Conclusions: Our study establishes the reference intervals for normal Chinese fetal lung biological parameters at 21–40 weeks. Moreover, the reference intervals apply to fetuses of different genders. This paper can provide a reference for the prenatal non-invasive assessment of fetal pulmonary hypoplasia. Full article

Three-dimensional echocardiography (3DE) is advised for right ventricular (RV) assessment. Data regarding the optimal acquisition settings and optimization are still scarce. We aimed to evaluate the feasibility, reproducibility and validation of 3DE for RV volume and function assessment, using cardiac magnetic resonance (CMR) as gold standard. Thirty healthy volunteers and 36 consecutive patients were prospectively included. CMR was performed in the latter. Standard apical four-chamber view (A4CV), focused A4CV and modified A4CV were used for 3DE RV acquisition. Feasibility (and the effect of changes in settings) was evaluated. Intra and interobserver analyses were performed by three observers (expert vs. novice). RV parameters by echocardiography were compared to CMR. Feasibility of acquisition was 16.7% for A4CV, 80.0% for focused A4CV and 16.7% for modified A4CV. Changes in settings had no significant influence on feasibility and further analysis. Intraobserver variability was good in both expert and novice, interobserver variability was good between experienced observers. Compared to CMR, 3DE volumes were significantly lower with fair to moderate correlation (EDV: 91.1 ± 24.4 mL vs. 144.3 ± 43.0 mL (p < 0.001), r = 0.653 and ESV: 48.1 ± 16.4 mL vs. 60.4 ± 21.2 mL (p < 0.001), r = 0.530, by multi-beat 3DE and CMR respectively). These findings suggest that standardization is needed in order to implement this technique in clinical practice, thus further studies are required. Full article

The application of segmentation methods to medical imaging has the potential to create novel diagnostic support models. With respect to fetal ultrasound, the thoracic wall is a key structure on the assessment of the chest region for examiners to recognize the relative orientation and size of structures inside the thorax, which are critical components in neonatal prognosis. In this study, to improve the segmentation performance of the thoracic wall in fetal ultrasound videos, we proposed a novel model-agnostic method using deep learning techniques: the Multi-Frame + Cylinder method (MFCY). The Multi-frame method (MF) uses time-series information of ultrasound videos, and the Cylinder method (CY) utilizes the shape of the thoracic wall. To evaluate the achieved improvement, we performed segmentation using five-fold cross-validation on 538 ultrasound frames in the four-chamber view (4CV) of 256 normal cases using U-net and DeepLabv3+. MFCY increased the mean values of the intersection over union (IoU) of thoracic wall segmentation from 0.448 to 0.493 for U-net and from 0.417 to 0.470 for DeepLabv3+. These results demonstrated that MFCY improved the segmentation performance of the thoracic wall in fetal ultrasound videos without altering the network structure. MFCY is expected to facilitate the development of diagnostic support models in fetal ultrasound by providing further accurate segmentation of the thoracic wall. Full article