Artificial Intelligence-Assisted Echocardiographic Image-Analysis for the Diagnosis of Fetal Congenital Heart Disease: A Systematic Review and Meta-Analysis

Background

To assess the precision of artificial intelligence (AI) in aiding the diagnostic process of congenital heart disease (CHD).

Methods

PubMed, Embase, Cochrane, and Web of Science databases were searched for clinical studies published in English up to March 2024. Studies using AI-assisted ultrasound for diagnosing CHD were included. To evaluate the quality of the studies included in the analysis, the Quality Assessment Tool for Diagnostic Accuracy Studies-2 scale was employed. The overall accuracy of AI-assisted imaging in the diagnosis of CHD was determined using Stata15.0 software. Subgroup analyses were conducted based on region and model architecture.

Results

The analysis encompassed a total of 7 studies, yielding 19 datasets. The combined sensitivity was 0.93 (95% confidence interval (CI): 0.88-0.96), and the specificity was 0.93 (95% CI: 0.88-0.96). The positive likelihood ratio was calculated as 13.0 (95% CI: 7.7-21.9), and the negative likelihood ratio was 0.08 (95% CI: 0.04-0.13). The diagnostic odds ratio was 171 (95% CI: 62-472). The summary receiver operating characteristic (SROC) curve analysis revealed an area under the curve of 0.98 (95% CI: 0.96-0.99). Subgroup analysis found that the ResNet and DenNet architecture models had better diagnostic performance than other models.

Conclusions

AI demonstrates considerable value in aiding the diagnostic process of CHD. However, further prospective studies are required to establish its utility in real-world clinical practice.

The PROSPERO registration

CRD42024540525, https://www.crd.york.ac.uk/prospero/display_record.php?RecordID=540525.

Novel approach to prenatal predictors of outcomes for fetuses with severe Ebstein anomaly

OBJECTIVE

Ebstein anomaly (EA) is a cardiac malformation with highly variable presentation and severity with limited perinatal management options. We present incorporation of fetal lung measurements into a multidisciplinary evaluation for counseling and predicting postnatal outcomes in patients with severe EA.

METHODS

Five fetuses with severe fetal EA were reviewed. Third trimester sonographic observed/expected total lung area (O/E TLA) and lung to head ratio (O/E LHR), fetal MRI total fetal lung volume ratio (O/E-TFLV), echocardiographic cardio-thoracic ratio (CT ratio), sonographic estimated fetal weight (EFW) by Hadlock formula and presence of hydrops, were used to guide perinatal management.

RESULTS

Three of five had appropriate fetal growth, were delivered at term in a cardiac operative suite, and underwent immediate intervention with good neonatal outcomes. Two had severe fetal growth restriction (FGR), CT ratios > 0.8 and O/E LHR and TLA < 25%. One of which delivered prematurely with neonatal demise and one suffered in utero demise at 34 weeks.

CONCLUSIONS

FGR, hydrops, increased CT ratio and reduced O/E LHR and TFLV are potential prognosticators of poor outcomes in severe EA, and should be validated in larger cohorts that would allow for a statistical analysis of the predictive utility of these measurements.

Utility of fetal cardiac magnetic resonance imaging in assessing the cardiac axis in fetuses with congenital heart disease

BACKGROUND

Fetal dedicated echocardiography is the standard to measure the fetal cardiac axis. However, fetal screening ultrasound (US) or fetal dedicated echocardiography may be technically limited.

OBJECTIVE

The purpose of this study was to explore the accuracy of fetal cardiac magnetic resonance imaging (MRI) to measure the cardiac axis in fetuses with congenital heart disease as an adjunct to fetal dedicated echocardiography and to assess the predictive value of fetal cardiac MRI measurements in distinguishing healthy fetuses from fetuses with congenital heart disease.

MATERIALS AND METHODS

This is a retrospective study of fetuses referred to our hospital for a fetal cardiac MRI from November 2019 to December 2021. Cardiac axes were measured in the 4-chamber view of the fetal heart using fetal cardiac MRI and dedicated echocardiography, or only using fetal cardiac MRI when screening US was technically limited. The fetuses were divided into a congenital heart disease group and a healthy control group. We used Bland-Altman analysis and the intraclass correlation coefficient (ICC) to assess the agreement of cardiac axis measurements in fetuses with congenital heart disease obtained by cardiac MRI and by fetal dedicated echocardiography. Receiver operating characteristic (ROC) curve analysis of the fetal cardiac axes in the congenital heart disease and healthy fetus groups assessed the predictive value of the cardiac axis measurements.

RESULTS

This retrospective study included 431 women (162 carrying fetuses with congenital heart disease, 269 carrying healthy fetuses). Cardiac axes were measured in the 162 fetuses with congenital heart disease using fetal cardiac MRI and dedicated echocardiography. Cardiac axes were measured in the 269 healthy control fetuses using fetal cardiac MRI when fetal screening US was technically limited. The interobserver analysis and intraobserver analysis showed that the cardiac axis measured by fetal cardiac MRI and fetal dedicated echocardiography was repeatable (ICC>0.90). In 162 fetuses with congenital heart disease, Bland-Altman analysis showed a strong agreement between cardiac MRI and fetal dedicated echocardiography measurements for the cardiac axis. The ICC for the cardiac axis values between cardiac MRI and fetal dedicated echocardiography measurements was 0.99. In fetuses with congenital heart disease, 64.2% (104/162) had an abnormal cardiac axis. For the fetal cardiac axis in both the 162 fetuses with congenital heart disease and the 269 healthy fetuses, the area under the ROC curve reached 0.85 (95% confidence interval: 0.80-0.89; P<0.0001).

CONCLUSION

The cardiac axis can be accurately measured using fetal cardiac MRI when fetal dedicated echocardiography/fetal screening US is technically limited. The cardiac axis measurements by fetal cardiac MRI are consistent with known cardiac axis measurements by fetal dedicated echocardiography. The frequency of abnormal cardiac axis depends on the type of congenital heart disease.

Deep Learning-Based Computer-Aided Fetal Echocardiography: Application to Heart Standard View Segmentation for Congenital Heart Defects Detection

Accurate segmentation of fetal heart in echocardiography images is essential for detecting the structural abnormalities such as congenital heart defects (CHDs). Due to the wide variations attributed to different factors, such as maternal obesity, abdominal scars, amniotic fluid volume, and great vessel connections, this process is still a challenging problem. CHDs detection with expertise in general are substandard; the accuracy of measurements remains highly dependent on humans' training, skills, and experience. To make such a process automatic, this study proposes deep learning-based computer-aided fetal heart echocardiography examinations with an instance segmentation approach, which inherently segments the four standard heart views and detects the defect simultaneously. We conducted several experiments with 1149 fetal heart images for predicting 24 objects, including four shapes of fetal heart standard views, 17 objects of heart-chambers in each view, and three cases of congenital heart defect. The result showed that the proposed model performed satisfactory performance for standard views segmentation, with a 79.97% intersection over union and 89.70% Dice coefficient similarity. It also performed well in the CHDs detection, with mean average precision around 98.30% for intra-patient variation and 82.42% for inter-patient variation. We believe that automatic segmentation and detection techniques could make an important contribution toward improving congenital heart disease diagnosis rates.

A 10-Year Retrospective Review of Prenatal Applications, Current Challenges and Future Prospects of Three-Dimensional Sonoangiography

Realistic reconstruction of angioarchitecture within the morphological landmark with three-dimensional sonoangiography (three-dimensional power Doppler; 3D PD) may augment standard prenatal ultrasound and Doppler assessments. This study aimed to (a) present a technical overview, (b) determine additional advantages, (c) identify current challenges, and (d) predict trajectories of 3D PD for prenatal assessments. PubMed and Scopus databases for the last decade were searched. Although 307 publications addressed our objectives, their heterogeneity was too broad for statistical analyses. Important findings are therefore presented in descriptive format and supplemented with the authors’ 3D PD images. Acquisition, analysis, and display techniques need to be personalized to improve the quality of flow-volume data. While 3D PD indices of the first-trimester placenta may improve the prediction of preeclampsia, research is needed to standardize the measurement protocol. In highly experienced hands, the unique 3D PD findings improve the diagnostic accuracy of placenta accreta spectrum. A lack of quality assurance is the central challenge to incorporating 3D PD in prenatal care. Machine learning may broaden clinical translations of prenatal 3D PD. Due to its operator dependency, 3D PD has low reproducibility. Until standardization and quality assurance protocols are established, its use as a stand-alone clinical or research tool cannot be recommended.

Description of misdiagnosis and missed diagnosis of fetal complex heart malformations by prenatal echocardiography combined with postnatal cardiovascular casting

OBJECTIVE

To compare prenatal echocardiography with postnatal cardiovascular casting for detection of fetal cardiovascular malformations, and to discuss the causes of prenatal misdiagnosis and missed diagnosis.

METHODS

We retrospectively identified patients from 2013 to 2018 at our Maternal-Fetal Medicine Center who were reported to have a fetal diagnosis of severe congenital heart malformations (CHMs). Subjects had postnatal confirmation of CHMs. Prenatal and postnatal medical records, including ultrasound results and casting findings, were reviewed and analyzed.

RESULTS

Postnatal casting showed that all 35 fetuses had complex CHMs. In these 35 cases, 90 cardiovascular malformations were found by postnatal casting, and 69 were detected by prenatal echocardiography. Among the other 21 cardiovascular malformations, 7 were misdiagnosed and 14 diagnoses were missed by prenatal ultrasound.

CONCLUSION

Prenatal echocardiography may lead to misdiagnosis and missed diagnoses, especially in cases with great arterial branching anomalies. Postnatal casting can demonstrate the configuration of the great vessels and smaller branches directly. Thus, understanding of such malformations via postnatal casting may help to improve prenatal diagnostic accuracy.

Radiologist's Guide to Diagnosis of Fetal Cardiac Anomalies on Prenatal Ultrasound Imaging

Congenital cardiac anomalies are a common finding during prenatal anatomical survey ultrasound examination. Cardiac anomalies are a major cause of prenatal and neonatal mortality and morbidity. If the anomaly is not lethal, most would require surgical correction. Therefore, early recognition of these abnormalities is essential for parental counseling and delivery planning, as well as analysis of neonatal treatment options. Although prenatal ultrasound plays an important role in identification of such anomalies, diagnosis and interpretation of imaging findings require familiarity and knowledge of the common imaging features. In this article, we provide a comprehensive review of ultrasound appearance of common fetal cardiac anomalies.