Results

There were 50 studies that broadly focused on the application of machine learning in paediatric cardiology research. These studies were categorised and focused on various intentions for clinical use (Figure 3): diagnosis and assessment of underlying critical and non-critical CHD (n = 20), ^{Reference Gharehbaghi, Dutoit, Sepehri, Kocharian and Lindén1,Reference Diller, Lammers and Babu-Narayan19–Reference Lv, Dong and Lei37} prediction and risk stratification of outcomes in CHD (n = 15), ^{Reference Ruiz, Saenz and Lopez-Magallon38–Reference Cainelli, Bisiacchi and Cogo52} management of patients with CHD (n = 2), ^{Reference Diller, Kempny and Babu-Narayan53,Reference Wolf, Lee and Nicolson54} medical device research (n = 4), ^{Reference Wang, Javadekar and Rajagopalan55–Reference Liu, Aslan and Hess58} novel genetics and biomarkers in CHD (n = 5), ^{Reference Liu, Zhao and Yuan59–Reference Qi, Zhang and Zhao63} CHD in pregnancy (n = 3), ^{Reference Ren, Zhu and Gao64–Reference Dozen, Komatsu and Sakai66} and social media research (n = 1). ^{Reference Klein, Sarker, Cai, Weissenbacher and Gonzalez-Hernandez67} The distribution of various machine learning algorithms used in these studies is summarised in Figure 4. The two most common algorithms were deep neural networks (deep learning) and support vector machines. Hidden Markov models and linear discriminant analysis were the least common algorithms.

Figure 3. Distribution of Machine Learning Applications and Uses in Pediatric Cardiology Research.

*Congenital Heart Defect (CHD).

Figure 4. Distribution of Machine Learning Algorithms Used in Pediatric Cardiology Research in General.

There were 20 studies focusing on the diagnosis and assessment of critical and non-critical CHD. The Table 1 summarises the details of these studies, including purpose, design, sample, machine learning technique, and primary findings. All studies were observational, generally of small sample size. None of the studies provided 100% of the Minimum Information About Clinical Artificial Intelligence Modeling checklist items (standard documentation guidelines). ^{Reference Norgeot, Quer and Beaulieu-Jones18} Included studies applied machine learning to auscultation of heart sounds in patients with CHD, interpreting transthoracic echocardiogram data, or processing medical images (cardiovascular MRI). As shown in Figure 5, deep neural networks and support vector machines were also the most commonly used classification algorithms in these studies. More importantly, Table 1 highlights that using cross validation on existing retrospective data, the overall accuracy of the various machine learning algorithms exceeded 80%, with some techniques reaching 95 to 100%. With the exception of two studies, none of these models were externally validated on independent data sets. The remainder of this scoping review will focus on the qualitative synthesis of 20 studies that focused on machine learning applications for the diagnosis and assessment of CHD.

Table 1. Summary of studies focusing on machine learning techniques applied to diagnosing and assessing CHD in neonates, infants, and children

* Area Under the Curve (AUC), Atrial Septal Defect (ASD), Atrioventricular (AV), Bicuspid Aortic Valve (BAV), Cardiovascular Magnetic Resonance Imagining (CMR), Coarctation of the aorta (CoA), Confidence Interval (CI), Congenital Heart Defects (CHD), Convolutional Neural Network (CNN), Deep Neural Network (DNN), Linear Discriminant Analysis (LDA), Interquartile Range (IQR), Left Ventricle (LV), Mitral Regurgitation (MR), Negative Predictive Value (NPV), Patent Ductus Arteriosus (PDA), Positive Predictive Value (PPV), Principal Component Analysis (PCA), Right Ventricle (RV), Root Mean-Square Error (RMSE), Signal-to-Noise Ratio (SNR), Standard Deviation (SD), Structural Similarity Index (SSIM), Support Vector Machine (SVM), Time Growing Neural Network (TGNN), Tricuspid Regurgitation (TR), Ventricular Septal Defect (VSD).

Figure 5. Distribution of Machine Learning Algorithms Used in Pediatric Cardiology Research Focusing on the Diagnosis and Assessment of Critical and Non-Critical CHD

The most common neural network types were the convolutional neural network, time growing neural network, autoencoder, and generative adversarial network.

The most common kernel among the studies who used support vector machine was a Gaussian kernel.

*Congenital Heart Defect (CHD).

Synthesis of literature

It is apparent that machine learning in paediatric cardiology research is an evolving field. Diagnosing and assessing patients with CHD can typically be done non-invasively, using examination findings and diagnostic tools based on auditory or visual pattern recognition. However, such medical images and signal data are considered unstructured; they are not stored as tabular data or in formatted fields. ^{Reference Kagiyama, Shrestha, Farjo and Sengupta68} These type of data historically require the time-consuming process of clinician review and interpretation. Thus, in individuals with CHD, machine learning techniques described in Table 1 have focused on accurate diagnosis and assessment based on the classification of auscultatory heart sounds, transthoracic echocardiograms, or cardiovascular magnetic resonance images.

Auscultation

Aortic valves are normally tricuspid in nature and provide an outlet for blood flow from the heart to the body. Bicuspid aortic valves occur in 0.5 to 2% of children. ^{Reference Gharehbaghi, Dutoit, Sepehri, Kocharian and Lindén1,Reference Mahle, Sutherland and Frias69–Reference Spaziani, Ballo and Favilli72} A subset of these patients develops progressive valve disease and/or aortic dilatation, with risk of life-threatening aortic aneurysm and dissection. Children with known bicuspid aortic valves must be monitored throughout their lives to measure risk for aortic aneurysm. Diagnosis of bicuspid aortic valves can be done via phonocardiogram, which records cyclical sounds produced by the heart. ^{Reference Gharehbaghi, Dutoit, Sepehri, Kocharian and Lindén1} Patients with a bicuspid aortic valve typically have a systolic ejection click. Identification of this click through traditional auscultation methods can be limited due to provider expertise and skill, as well as rapid heart rates in young children. Gharehbaghi and colleagues performed a study that collected phonocardiogram data prospectively. ^{Reference Gharehbaghi, Dutoit, Sepehri, Kocharian and Lindén1} This study created a statistical time growing neural network that automatically classified bicuspid aortic valves through use of recorded heart sounds produced by the phonocardiogram. ^{Reference Gharehbaghi, Dutoit, Sepehri, Kocharian and Lindén1} The phonocardiogram recordings were preprocessed, so that the cardiac cycle could be segmented to recognise the additional heart sound (systolic ejection click). This segmentation was used to build a classifier to identify healthy subjects and those with a bicuspid aortic valve. The model was able to classify the 865 cardiac cycles with 98.5% accuracy. This improved diagnostic sensitivity for a subtle exam finding stands to bring many previously unrecognised children and adults to appropriate cardiac care.

Several other investigators are also using non-invasively recorded heart sounds to diagnose cardiac disease. ^{Reference Aziz, Khan, Alhaisoni, Akram and Altaf29,Reference Gharehbaghi, Sepehri and Babic30,Reference Gómez-Quintana, Schwarz and Shelevytsky34} For example, Elgendi and colleagues ^{Reference Elgendi, Bobhate and Jain27} used linear discriminant analysis to detect pulmonary arterial hypertension using digital auscultation to record the unique vibrations of the hypertensive pulmonary circulation. Their model performed well, with a sensitivity of 84% and specificity of 88.6% for entropy (disorder of heart sound pattern) of the first sinusoid formant (frequency resonance of heart sounds). Sun and colleagues, ^{Reference Sun and Wang20} aimed to diagnose small, medium, and large ventricular septal defects based on heart sound feature extraction, using classification boundary curves and ellipse models. The ellipse model outperformed five other models used in the study for normal, small, medium, and large ventricular septal defect classification (accuracy 99%, 95.5%, 92.1%, and 96.2%). Cardiac auscultation is nuanced, but clearly machine learning holds promise for improving diagnostic accuracy for providers at all experience levels.

Transthoracic echocardiogram

Echocardiography is the mainstay of non-invasive assessment for CHD, but requires experienced reviewers that remain susceptible to biases. Coarctation of the aorta, a common form of critical CHD characterised by narrowing of the thoracic aorta, is particularly well suited for echocardiographic diagnosis, though poses challenges when it comes to image interpretation. ^{Reference Pereira, Bueno and Rodriguez23} Narrowing of the aorta causes obstruction of normal blood flow to the body and excessive pressure to be generated by the left ventricle. If the obstruction is not diagnosed in a timely manner, it can lead to heart failure and poor systemic perfusion. Pereira and colleagues ^{Reference Pereira, Bueno and Rodriguez23} retrospectively collected 2-dimensional echocardiographic images of the aortic arch. The aim of this study was to develop a fully automated algorithm to detect coarctation of the aorta from 2-dimensional echocardiographic images using standard view planes (suprasternal, apical, and parasternal windows). Static images representing a single cardiac cycle (end diastolic and end systolic phases) were pre-selected for model development. Neonates born with coarctation of the aorta alone and healthy neonates born without coarctation of the aorta were included in the sample. A stacked denoising autoencoder neural network was used for feature extraction over predefined image regions (sectors). A support vector machine classifier was trained on a random subset of training data features. The parasternal long axis view had the lowest coarctation error rate (end diastolic phase [7.7], end systolic phase [11.5]), and the apical view had the lowest healthy error rate (end diastolic phase [20.0], end systolic phase [20.0]). ^{Reference Pereira, Bueno and Rodriguez23} When the views were combined, more undecided cases resulted. The increase in undecided cases with inclusion of multiple views is not surprising given the increased model complexity, and therefore difficulty for the model to classify outcomes.

Further research has focused on image denoising, automatic detection, and clustering subjects based on quantitative image data to further support clinical decision making and diagnosis of critical and non-critical CHD. ^{Reference Wang, Liu and Wang33} Diller et al. ^{Reference Diller, Lammers and Babu-Narayan19} aimed to remove acoustic shadowing artifacts that occur during transthoracic echocardiograms through the use of a deep neural network and autoencoder. Cross-entropy (a loss function that measures differences between two probability distributions -- original image vs. reconstructed image) ^{Reference Diller, Lammers and Babu-Narayan19,Reference Rubinstein73} and sum of squared differences (measures image quality) were performance evaluation metrics used on the test data set. Autoencoders extracted features and trained significantly better on CHD samples compared to healthy samples, represented by a lower cross-entropy and lower mean squared difference (0.2597 ± 0.0327 and 118.86 ± 61.52). ^{Reference Diller, Lammers and Babu-Narayan19} Finally, Meza et al. ^{Reference Meza, Slieker and Blackstone21} used an unsupervised hierarchical cluster analysis in 651 neonates with critical left heart obstruction to determine if subjects could be defined using more clinically meaningful clusters. Images were derived from transthoracic echocardiograms, and three distinct groups emerged (n = 215, 338, 98). Aortic valve atresia and left ventricular end diastolic volume variables significantly distinguished the groups. Median left ventricular end diastolic area for groups 1, 2, 3 was 1.35, 0.69, and 2.47 cm² (p < 0.0001). Aortic atresia in groups 1, 2, 3 was present in 11%, 87%, and 8%, (p < 0.0001). ^{Reference Meza, Slieker and Blackstone21} The authors suggest that clustering analyses yield more reliable delineation of subject heart structure characteristics. These data support the use of clustering approaches to more accurately diagnose CHD. The added complexity and volume of data in diagnostic imaging poses a challenge to machine learning, but these investigators demonstrate the potential value machine learning brings to clinical pattern fitting and decision making.

Cardiovascular MRI

Cardiovascular MRI is considered the clinical gold standard for accurate assessment of ventricular volumes and function. Obtaining quality images requires repetitive breath holding, which can be a challenge for patients with CHD who may suffer from shortness of breath at baseline or who are simply too young to comply with breath-holding instructions. The investigators of the next study explored alternative ways to denoise cardiovascular magnetic resonance images that were captured during free-breathing. Specifically, they aimed to use real-time imaging, while applying reconstruction techniques to denoise the images (artifact versus artifact free images). ^{Reference Hauptmann, Arridge, Lucka, Muthurangu and Steeden28} Metrics used to evaluate the performance of their convolutional neural network were the signal-to-noise ratio, acceleration factor, and image cropping. The root-mean square error and the structural similarity index were used to evaluate the test data set reconstructed image accuracy. The continuously rotating tiny golden angle sampling pattern had the lowest root-mean square error, and the highest structural similarity index (p < 0.0001) compared to all other sampling methods. The signal-to-noise ratio decreased from 20 dB to 10 dB, and the acceleration factor increased from 10x to 16x. Finally, the reconstruction time for all slices originating from raw data was 5.6x faster for the convolutional neural network (22 seconds). ^{Reference Hauptmann, Arridge, Lucka, Muthurangu and Steeden28} This study demonstrates that machine learning models have the potential to successfully remove artifact, while decreasing reconstruction times to produce better quality images and more accurate measurements in real-time. More importantly, if these models become widely used, they have the potential to improve patient comfort or decrease the need for sedation in infants and younger children during cardiovascular MRI, as the need for frequent breath holding is no longer required.

Further, clustering subjects based on their cardiovascular MRI data is a very popular approach presented in the literature. While cardiac imaging has already been established as a standard and definitive way to diagnose complex heart disease, investigators are now taking advantage of the machine learning applications that can unveil hidden patterns in these data to boost diagnostic accuracy. ^{Reference Karimi-Bidhendi, Arafati, Cheng, Wu, Kheradvar and Jafarkhani31,Reference Lu, Fu, Li and Qi32,Reference Tandon, Mohan and Jensen35} Bruse and colleagues ^{Reference Bruse, Zuluaga and Khushnood24} used agglomerative hierarchical clustering and principal component analysis to detect clinically meaningful shape clusters using anatomical cardiovascular MRI data. Subjects with and without surgically corrected coarctation of the aorta and subjects with healthy aortic arches were included. For each cross-validated run, 83% of healthy aortic arches were assigned to the healthy group, 85% of the coarctation shapes were correctly assigned, and 100% of the surgically corrected shapes were accurately assigned. Human clinicians would ideally use the results provided by the machine learning model to either compare or validate their own interpretation of the image data. This suggests that clustering techniques have the potential to inform clinical decision making at the time of diagnosis, thus improving accuracy and efficiency.

Finally, cardiac segmentation applied to cardiovascular magnetic resonance images is a process that takes a complex multidimensional image of the heart and separates its major sections, for example, the ventricles and coronary arteries. Segmentation is necessary because each ventricle or vessel can then be assessed quantitatively. Particularly, ventricular mass, volume, or ejection fraction can be quantitively measured. Segmentation driven by deep learning; specifically convolutional neural networks require large amounts of training data. Research studies aiming to optimise cardiac segmentation driven by deep learning in patients with CHD are faced with data accessibility challenges related to the extreme heterogeneity of cardiac anatomy and rarity of disease within the CHD population. Generative adversarial networks learn from real images in order to generate synthetic image data. The generative adversarial network has two networks that compete with one another. A generator network creates false images that the discriminator network will use to decipher between real and false images. ^{Reference Chen, Qin and Qiu74} Through these networks, new synthetic images are produced and can increase the size of training data sets in populations with limited image data. ^{Reference Zeleznik, Weiss and Taron75} Investigators used these methods to create synthetic image data in patients with Tetralogy of Fallot, an extremely rare heart condition in the general population but one of the most common critical CHD, and found the images to be anatomically accurate. ^{Reference Diller, Vahle and Radke76} Generative adversarial networks may have major implications in CHD diagnostic image research, as they can potentially expand training data sets and allow models to predict rare and life-threatening diseases.