Detection Of Heart Defects Research Articles

Literature Review of the Use of Deep Learning Methods in Identification of Heart Defects Based on ECG

In recent years the of Deep Learning has been widely used in various fields of science. In the field of health many studies use deep learning, especially related to heart disorders. Identification and detection of heart defects that are widely used are by using an electrocardiogram (ECG). This article aims to explain the trend of ECG research using deep learning approaches in recent years. We reviewed journals with the keyword title "ECG Deep Learning" and published from 2016 to October 2023. The articles that have been obtained are then classified based on the most frequently discussed topics including: data sets, pre-processing, feature extraction, and classification/identification methods. Research related to ECG and the application of deep learning methods is very interesting for researchers, because researchers trace that there are still few studies that focus on researching related to it. This article will provide further explanation of the most widely used algorithms for ECG research with a deep learning approach. In this article, critical aspects of ECG research can be carried out in the future, namely the use of data in the form of other data from ECG, and the use of deep learning is a very big opportunity for researchers.

The Value of Fetal Echocardiography for Early Detection of Congenital Heart Defects: A Report of our Experience in a Great Referral Heart Center in Iran

Background: Prenatal diagnosis of cardiac defects in fetus is of paramount importance to enable a better monitoring of heart defects and to allow planning a proper management. We reported our experience on fetal heart echocardiography in a great referral heart center in Iran with the approach to assessment of the diagnostic value of this tool for detecting Congenital Heart Diseases (CHDs). Methods: This retrospective cohort study included 234 pregnant women attended at a great referral heart center in Tehran, Iran between 2012 and 2016 for heart abnormalities screening. All women were evaluated by trans-abdominal fetal echocardiography from 12 weeks of gestation to term and after birth. Results: Fetal echocardiography showed different types of CHDs in 5.9%. CHDs detected in 9.9% of neonates by echocardiography after birth. The most common heart abnormalities detected was VSD in 4.3%. Considering echocardiography findings after birth as the reference, fetal echocardiography had a sensitivity of 92.8%, a specificity of 66.6%, a positive predictive value of 59.8%, a negative predictive value of 86.6%, and an accuracy of 84.1% for detection of heart defects. Conclusion: Fetal heart echocardiography is a sensitive tool for detection of CHDs, but its specificity is moderate.

APPLICATION OF ARTIFICIAL NEURAL NETWORKS IN THE IDENTIFICATION OF HEART ABNORMALITIES BASED ON ECG: LITERATURE REVIEW

In recent years, a lot of research on Biomedicine is mainly related to heart defects. The recognition and detection of heart defects using an electrocardiogram (ECG) is numerous. This article aims to explain ECG research trends using the artificial neural networks (ANN) approach in the last twelve years. We reviewed journals with the keyword title "ANN ECG" and published from 2011 to 2022. Articles classified by the most frequently discussed topics include: data sets, case studies, pre-processing, feature extraction, and classification/identification methods. Data collection from articles obtained some of them use data banks such as PhysioNet, European Society of Cardiology (ESC), UCI Repository, as well as data collection directly from patients. The pre-processing stage of the total number of journals used 23% uses this stage, while in characteristic extraction almost 25% of papers are used. This research is very interesting because only a few researchers focus on researching about it. This article will provide further explanation of the most widely used algorithms for ECG research with the ANN approach. At the end of this article, critical aspects of ECG research can be carried out in the future, and the use of deep learning becomes a huge opportunity for researchers.

A phased approach for the adaptation of telemedicine in arrythmia management.

Cardiovascular disease is the world's leading cause of death. People at higher risk of heart disease need early detection to prevent or delay serious disease complications. In this paper, we present a new design of a wireless portable ECG (electrocardiogram). The ECG signal was preprocessed by analog filters and amplifiers before it was converted into digital data via an Arduino board, a highly available device for real-time signal processing. In order to allow communication between Arduino and Android phone, we use a Bluetooth module. The proposed mobile telemedicine technique allows immediate and rapid detection of heart defects followed by an alarm message from the patient to the doctor.

Phonocardiogram Classification Based on Machine Learning with Multiple Sound Features

In this study the heartbeat sound signals were tackled by classifying them into heart disease categories such as normal, artifact, murmur and extrahals in an attempt for early detection of heart defects. Phonocardiogram (i.e., PCG) is used to obtain the digital recording dataset of the heart sounds using an electronic stethoscope or mobile device. Multiple features are extracted from the digital recording dataset such as MFCC, Delta MFCC, FBANK and a combination between MFCC and FBANK features. Moreover, to classify the heartbeat sound signals, multiple well-known machine learning classifiers were used such as Naive Bays (NB), Support Vector Machine (SVM), Decision Tree (DT), Random Forest (RF), K-Nearest Neighbor (KNN) and Artificial Neural Network (ANN). The evaluation processes went through five metrics: Confusion matrix, accuracy, F1 score, precision and recall evaluating the recognition rate. Comparative experimental results show that the correctness of the feature with a best accuracy 99.2% adopted by MFCC and FBANK combination features which reduce false detection.

Automated detection of heart defects in athletes based on electrocardiography and artificial neural network

AbstractElectrocardiography (ECG) has proven to be one of the most efficient ways of tracking heart defects in athletes. However, the interpretation of electrocardiograms often require the expertise of a cardiologist. Meanwhile, an automated heart monitoring system could be used to ensure early heart defect detection in athletes, even in the absence of a cardiologist. In this paper, an automated heart defect detection model is proposed for athletes using ECG and Artificial Neural Network (ANN). We developed an ECG biomedical equipment to acquire 400 ECG data vectors from 40 participants, who comprises of athletes and non-athletes. Four classes of possible heart conditions among athletes, namely: normal, tachyarrhythmia, bradyarrhythmia and hypertrophic cardiomyopathy were considered. The ECG data collected were pre-processed and features were extracted based on first order moment. Different ANNs were trained to correctly classify the ECG data. By and large, the performances of ANNs that were trained based...

Screening performances of abnormal first-trimester ductus venosus blood flow and increased nuchal translucency thickness in detection of major heart defects.

The aim of this study was to evaluate the screening performances of abnormal ductus venosus (DV) blood flow for the detection of heart defects in chromosomally normal fetuses with increased nuchal translucency (NT) thickness at 11-13 + 6 weeks' gestational in a population of singleton pregnancies. During an 8-year period, all singleton pregnancies from 11 + 0 to 13 + 6 weeks were scanned for NT and DV blood flow assessment. Two groups of cases with abnormal NT were evaluated: NT ≥ 95th and NT ≥ 99th centile. DV waveforms were considered to be abnormal if the a-wave was reversed or absent (R/A). Addition of DV R/A a-wave to either NT ≥ 95th or NT ≥ 99th percentile increased specificity (p < 0.001 and p < 0.001, respectively), but not screening performances in detection of major heart defects (p = 0.73 and p = 0.91, respectively). Combination of DV R/A a-wave with NT ≥ 95th or NT ≥ 99th centile correlated with right heart defects (p = 0.024 and p = 0.013, respectively). In chromosomally normal fetuses, addition of abnormal DV a-wave to increased NT does not improve screening performances of NT in detection of major hearts defects in first trimester. However, there is correlation of such parameter with right heart defects and AV septal defects.

FETAL CARDIAC DISEASE: THE PATHOLOGIST'S PERSPECTIVE

In the European Union, it is estimated that 36,000 children are born every year with congenital heart disease and that a further 3000 who are diagnosed with congenital heart disease die as a result of termination of pregnancy, late fetal death or early neonatal death. In a normal population, the risk of a woman having a child with a congenital heart malformation is of the order of 0.8–1%, the risk rising to 2–3%, if a previous pregnancy was affected by heart disease and approaching 6% if the mother herself has a congenital heart defect. There is great variation between countries in the antenatal detection of heart defects, being lowest in those countries without ultrasound antenatal screening programmes (8–11%), but in Western Europe the detection rates vary between 19% and 48%. High-resolution echocardiography enables assessment of precise structures during the second trimester or even earlier.

Critical heart defects—the diagnostic challenge

Infants with potentially life threatening congenital heart defects (CHDs) are discharged from hospital after birth with the condition unrecognized. Improved prenatal ultrasound imaging and universal pulse oximetry screening of babies in nurseries are strategies that probably most would contribute to avoid such defects to be missed. In general combining first day of life pulse oximetry, clinical examination and echocardiography before discharge in suspect cases is a rational strategy for early postnatal detection of heart defects. Universal echocardiography screening of newborns may be too resource consuming to be cost-effective.

Prenatal detection of heart defects in a non-selected population of 30 149 fetuses–detection rates and outcome

Prenatal detection of heart defects in a non-selected population of 30 149 fetuses–detection rates and outcome

Prenatal detection of heart defects in a non‐selected population of 30 149 fetuses—detection rates and outcome

To evaluate the detection rate of congenital heart defects (CHD) in a non-selected population and to follow outcome after diagnosis. All 30,149 fetuses/newborns that were scheduled to deliver at our hospital between February 1991 and December 2001 were registered prospectively. Of these, 29,460 (98%) fetuses had a prenatal ultrasound scan at our center. The routine fetal examination at approximately 18 weeks' gestation included the four-chamber view and the great arteries of the fetal heart. The follow-up period was 2-13 years. Of 97 major CHDs, 55 (57%) were detected prenatally, 16% (9/55) prior to, 66% (36/55) at and 18% (10/55) after the routine scan. Forty-four percent (19/43) of the isolated CHDs, 67% (36/54) of those with associated malformations and 48% (11/23) of the isolated ductal-dependent CHDs were detected. Thirty-eight percent (37/97) had an abnormal karyotype. Of the 55 major CHDs detected, 44% (24) of the pregnancies with lethal/serious fetal malformations were terminated, 15% (8) died in utero, 42% (23) were born alive and 27% (15) were still alive after 2 years. Of the 42 CHDs detected postnatally, 2% (1) were terminated for other reasons, 98% (41) were born alive and 81% (34) were still alive after 2 years. Prenatal detection of CHD is still a challenge, with a 57% detection rate only. Isolated defects are detected less frequently. The overall outcome suggests that the most severe defects are detected with the present screening setting; only 27% of the babies with major CHDs detected were still alive after 2 years. Data from long-term follow-up will be of importance for the counseling process.

Cardiac screening examination of the fetus: guidelines for performing the ‘basic’ and ‘extended basic’ cardiac scan

Cardiac screening examination of the fetus: guidelines for performing the ‘basic’ and ‘extended basic’ cardiac scan

Antenatal Detection of Heart Defects is Important and Achievable

Congenital heart disease occurs in about 1 in 125 live births, approximately half of which require urgent evaluation and management. Cardiovascular malformations account for 10% of infant deaths and nearly half of all deaths from all malformations. They are, therefore, both important and common, but are still frequently missed in primary screening programmes worldwide. The UK average detection rate is 23% with wide postcode variation. The reasons for missing defects are difficult to ascertain. However, with 'hands-on' training of obstetric sonographers in their own unit, detection levels may increase to 70–80%. With ongoing support these levels are sustainable. This 'hands-on' training uses a systematic protocol, the five transverse views, which allows the majority of defects to be detected at the routine 20-week scan without increasing routine examination time.

The ‘spin’ technique: a new method for examination of the fetal outflow tracts using three‐dimensional ultrasound

The prenatal detection of congenital heart defects remains one of the most difficult challenges for the sonologist/sonographer when performing the second- or third-trimester screening examination. The four-chamber view has been used for a number of years as the primary screening image for detection of heart defects, but the inclusion of the right and left outflow tracts increases the detection of cardiac malformations. One of the difficulties, however, is obtaining and interpreting two-dimensional images of the outflow tracts. This paper reviews a new technique using three-dimensional (3D) multiplanar imaging that allows the examiner to identify the outflow tracts within a few minutes of acquiring the 3D volume dataset by rotating the volume dataset around the x- and y-axes. 3D multiplanar imaging of the fetal heart using static 3D or spatio-temporal image correlation (STIC) imaging allows the examiner to obtain a volume of data that can be manipulated along the x- and y-axes using reference points from the four-chamber view, five-chamber view, three-vessel view at the level of the bifurcation of the pulmonary arteries, and three-vessel view at the level of the transverse aortic arch and trachea. The full length of the main pulmonary artery, ductus arteriosus, aortic arch and superior vena cava could be identified easily in the normal fetus by rotating the volume dataset along the x- and y-axes. The vessels were identified using the four-chamber view, the five-chamber view, and the two three-vessel views. The technique was useful in identification of d-transposition of the great vessels and evaluation of the outflow tracts in hypoplastic left heart syndrome. 3D multiplanar evaluation of the fetal heart allows the examiner to identify the outflow tracts using a simple technique that requires only rotation around x- and y-axes from reference images obtained in a transverse sweep through the fetal chest.

Cardiac malformations in first-trimester fetuses with increased nuchal translucency: ultrasound diagnosis and postmortem morphology.

The aim of this study was to explore the diagnostic accuracy of first-trimester transvaginal echocardiography in fetuses with increased nuchal translucency (NT) thickness, by comparing the ultrasound diagnosis with the findings on postmortem examination or mid-gestational ultrasound and neonatal outcome. Transvaginal echocardiography was performed in 45 fetuses with a NT > 95th centile. Karyotyping was performed in 43. In 20 of the 23 pregnancies in which termination of pregnancy was carried out, postmortem examination was performed to determine the presence and type of heart defect. Mid-gestational echocardiography was performed in ongoing pregnancies and neonatal follow-up information was obtained. Findings on first-trimester transvaginal echocardiography were compared to those of second-trimester echocardiography or the results of postmortem examination. The mean NT in the fetuses with and without heart defects was calculated. Of the 45 fetuses, heart malformations were suspected on first-trimester ultrasound in 10 (22%), of which eight (80%) were found to have a chromosomal abnormality. Postmortem examination showed minor additional findings in some cases and major discrepancies occurred in none. Septal defects were the most common defects in trisomic fetuses. In three fetuses with a 45 X karyotype, hypoplastic left heart syndrome was diagnosed. Heart defects were diagnosed in three euploid fetuses in which fetal demise occurred. The sensitivity and specificity for the detection of heart defects of transvaginal echocardiography were 88% and 97%, respectively. The mean NT in fetuses with a normal heart (4.3 mm) was significantly smaller than that of fetuses with heart defects (7.4 mm). Transvaginal echocardiography can be performed reliably in first-trimester fetuses with an increased NT. In this study, the proportion of chromosomally abnormal fetuses with a heart defect was not different from that found in newborns, except for cases of Turner syndrome. Fetal demise occurred in all three euploid fetuses with a heart malformation. The fetuses with a heart defect had a larger NT than did those without.

Prenatal detection of heart defects at the routine fetal examination at 18 weeks in a non‐selected population

Few studies have addressed the prenatal detection rate of congenital heart defects in a non-selected population at 18 weeks of gestation. Our objective was to assess the change in the prenatal detection rate of congenital heart defects in such a population that resulted from incorporating the four-chamber view at the second-trimester routine ultrasound examination. The prenatal detection rate of heart defects was prospectively compared between 4435 fetuses in Phase I who were scanned without special attention to the heart, and 7459 fetuses in Phase II who were scanned incorporating the four-chamber view. Of the 49 heart defects in Phase 1, 17 (35%) were critical and three (18%) of these were detected prenatally. Of the 90 heart defects in Phase II, 23 (26%) were critical, six (26%) of these were detected prenatally at the 18 weeks' routine scan, and three were detected in the third trimester, providing a total prenatal detection rate of 39%. A defect was classified as critical when a surgical repair was likely to be required because of gross structural complexity having a functional significance, e.g. transposition of the great arteries, hypoplastic left heart syndrome, atrioventricular septal defect, coarctation of the aorta, and large ventricular septal defect. No non-critical heart defects were detected prenatally in either of the phases. The incidences in the total population were 11 and 12/1000 in Phases I and II, respectively. Thirty-two per cent of the critical and 16% of the non-critical defects had associated abnormalities and/or abnormal karyotype.(ABSTRACT TRUNCATED AT 250 WORDS)