A classifier based on the artificial neural network approach for cardiologic auscultation in pediatrics

Bhatikar, Sanjay; DeGroff, Curt; Mahajan, Roop L.

doi:10.1016/j.artmed.2004.07.008

Cited by 103 publications

(52 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Several groups have applied a variety of signal processing techniques to pediatric heart sound recordings and report sensitivity and specificity values approaching 100%. [120][121][122][123][124] Automatically generated computer-aided auscultation interpretation can also provide immediate decision support to the primary care provider. 125 With computer-aided auscultation used this way, the referral sensitivity and specificity of primary care providers increased, with the potential to significantly reduce unnecessary and costly referrals.…”

Section: Tele-auscultationmentioning

confidence: 99%

Telemedicine in Pediatric Cardiology: A Scientific Statement From the American Heart Association

Satou¹,

Rheuban²,

Alverson³

et al. 2017

Circulation

View full text Add to dashboard Cite

A dvancements in technology and broadband have revolutionized the current practice of medicine. The field of pediatric cardiology is no exception given the need for prompt diagnosis and reliance on cardiac imaging to identify infants and children with potentially life-threatening cardiovascular disease. As the relationship between telemedicine and pediatric cardiology has advanced, it has created a need to develop a broad, comprehensive document reviewing all the various aspects of telemedicine in pediatric cardiology. For more than a decade, a significant body of literature has been published describing individual experiences and practices, yet there remains no comprehensive statement or document summarizing this rapidly advancing field. In an effort to describe the collective experience and to provide structure and guidance for pediatric cardiology practitioners and healthcare providers, we have developed a scientific statement on the use of telemedicine in pediatric cardiology.Specific areas explored in this document include both neonatal and fetal teleechocardiography, implications for training community sonographers, pulse oximetry programs, qualitative improvement and appropriate use criteria initiatives, and remote electrophysiological monitoring. This document also includes teleconsultation and teleausculation, direct-to-consumer and home monitoring programs, and a look into the use of telemedicine and pediatric cardiology in the intensive care setting. Furthermore, a detailed review of the legislative, public policy, and legal aspects of telemedicine is provided, along with financial and reimbursement information.Several terms are used in the literature interchangeably; a brief explanation is provided to help readers of this document. The term telehealth is defined as the use of technology to bridge distances in any aspect of medicine; telemedicine is the specific application of technology to conduct clinical medicine at a distance. The term telecardiology is defined as the broad application of telemedicine in the field of cardiology specifically, and tele-echocardiography is the most common application used within this field. ECHOCARDIOGRAPHY AND TELEMEDICINEEchocardiography is the most commonly used noninvasive cardiovascular imaging modality and is considered to be both safe and cost-effective. Tele-echocardiography can be described as a process in which a provider or a technician obtains cardiovascular ultrasound images from a given patient and these images are subsequently transmitted to an offsite location where a cardiologist can provide further analysis and interpretation. Thus, tele-echocardiography enables expert interpretation and consultation in a rapid and potentially geographically disparate fashion, enabling prompt and accurate decision making involving triage, transport, and therapeutic priorities. Tele-echocardiography is now routinely used across the age and subspecialty spectrum in pediatric cardiology.

show abstract

Section: Tele-auscultationmentioning

confidence: 99%

Telemedicine in Pediatric Cardiology: A Scientific Statement From the American Heart Association

Satou¹,

Rheuban²,

Alverson³

et al. 2017

Circulation

View full text Add to dashboard Cite

show abstract

Section: Discussionmentioning

confidence: 99%

“…Different support-decision systems are developed using phonocardiography and automatic classifiers (37). An ANN-based classifier can be used for heart sound analysis, where different approaches may precede the classification process, such as the previously described wavelet representation (37). The phonocardiograms (38) were subjected to a fast Fourier transform to extract the energy spectrum in the frequency domain to detect heart murmurs in children.…”

Section: Discussionmentioning

confidence: 99%

Phonocardiography-based mitral valve prolapse detection using an artificial neural network

Bogdanović¹,

Ivan²,

Gavrovska³

et al. 2013

Ser J Exp Clin Res

View full text Add to dashboard Cite

“…Several groups have addressed the problem of frequency-domain classification of the phonocardiogram (PCG). In Bhatikar et al [3], an artificial neural network (ANN) is used to differentiate between innocent and pathological murmurs based on spectral information. In their study, manual selection and segmentation of individual heart cycles of acceptable quality is performed.…”

Section: Introductionmentioning

confidence: 99%

Nonnegative Matrix Factorization and Random Forest for Classification of Heart Sound Recordings in the Spectral Domain

Antink

Becker

Leonhardt

et al. 2016

2016 Computing in Cardiology Conference (CinC)

View full text Add to dashboard Cite

Stimulating the development of robust algorithms for the automated classification of phonocardiograms (PCGs) is the goal of the PhysioNet/CinC challenge 2016. In this paper, an approach to classify PCGs in the spectral domain is presented. First, the magnitude spectrogram is calculated. Next, the spectral shapes of four states of the cardiac cycle ("S 1 ","Systole", "S 2 ", "Diastole") are extracted using nonnegative matrix factorization, which is initialized with a time-domain segmentation algorithm. A Random Forest with 3000 trees is used for classification. Using 10-fold cross-validation on the unbalanced training data, a mean sensitivity of 0.92 at a specificity of 0.83 was achieved, resulting in an overall score of 0.88. On the complete hidden test data, a top score of 0.78 during phase II of the challenge with a sensitivity of 0.74 and a specificity of 0.83 was achieved.

show abstract

A classifier based on the artificial neural network approach for cardiologic auscultation in pediatrics

Cited by 103 publications

References 17 publications

Telemedicine in Pediatric Cardiology: A Scientific Statement From the American Heart Association

Telemedicine in Pediatric Cardiology: A Scientific Statement From the American Heart Association

Phonocardiography-based mitral valve prolapse detection using an artificial neural network

Nonnegative Matrix Factorization and Random Forest for Classification of Heart Sound Recordings in the Spectral Domain

Contact Info

Product

Resources

About