Artificial Intelligence-Assisted Auscultation of Heart Murmurs: Validation by Virtual Clinical Trial.

Artificial intelligence (AI) has potential to improve the accuracy of screening for valvular and congenital heart disease by auscultation. However, despite recent advances in signal processing and classification algorithms focused on heart sounds, clinical acceptance of this technology has been limited, in part due to lack of objective performance data. We hypothesized that a heart murmur detection algorithm could be quantitatively and objectively evaluated by virtual clinical trial. All cases from the Johns Hopkins Cardiac Auscultatory Recording Database (CARD) with either a pathologic murmur, an innocent murmur or no murmur were selected. The test algorithm, developed independently of CARD, analyzed each recording using an automated batch processing protocol. 3180 heart sound recordings from 603 outpatient visits were selected from CARD. Algorithm estimation of heart rate was similar to gold standard. Sensitivity and specificity for detection of pathologic cases were 93% (CI 90-95%) and 81% (CI 75-85%), respectively, with accuracy 88% (CI 85-91%). Performance varied according to algorithm certainty measure, age of patient, heart rate, murmur intensity, location of recording on the chest and pathologic diagnosis. This is the first reported comprehensive and objective evaluation of an AI-based murmur detection algorithm to our knowledge. The test algorithm performed well in this virtual clinical trial. This strategy can be used to efficiently compare performance of other algorithms against the same dataset and improve understanding of the potential clinical usefulness of AI-assisted auscultation.

Computerized Automatic Diagnosis of Innocent and Pathologic Murmurs in Pediatrics: A Pilot Study.

OBJECTIVE: Computer-aided auscultation in the differentiation of pathologic (AHA class I) from no or innocent murmurs (AHA class III) would be of great value to the general practitioner. This would allow objective screening for structural heart disease, standardized documentation of auscultation findings, and may avoid unnecessary referrals to pediatric cardiologists. Our goal was to assess the quality of a novel computerized algorithm that automatically classifies murmurs in phonocardiograms (PCGs) acquired in a pediatric population. DESIGN: This is a pilot study testing the ability of a novel computerized algorithm to accurately diagnose PCGs compared with interpreted echocardiograms as a gold standard. SETTING: This study was performed in pediatric cardiology clinics at a tertiary care hospital. PATIENTS: All incoming patients were recruited, including patients with no murmurs, innocent murmurs, and pathologic murmurs (106 patients). INTERVENTION: Using an electronic stethoscope, PCGs were acquired by the pediatric cardiologist from each patient. The PCGs were analyzed by the algorithm and diagnoses were compared with findings by echocardiograms interpreted by pediatric cardiologists which were used as the gold standard. OUTCOME MEASURES: Sensitivity, specificity, positive predictive value, negative predictive value, and accuracy were calculated. RESULTS: When compared with echocardiography as a gold standard in diagnosing murmurs, the computerized algorithm tested on N=34 PCGs, yielded a sensitivity of 87% and specificity of 100%, a positive predictive value of 100%, negative predictive value of 90% and an accuracy of 94%. CONCLUSION: With echocardiogram as a gold standard, this computerized algorithm can detect pathologic murmurs with high sensitivity, specificity and accuracy, comparable to if not better than published results of pediatric cardiologists and neonatologists. This study confirms the high quality and "real-world" robustness of a novel computational algorithm in the assessment of pediatric murmurs.

Modelling non-symmetric collagen fibre dispersion in arterial walls.

New experimental results on collagen fibre dispersion in human arterial layers have shown that the dispersion in the tangential plane is more significant than that out of plane. A rotationally symmetric dispersion model is not able to capture this distinction. For this reason, we introduce a new non-symmetric dispersion model, based on the bivariate von Mises distribution, which is used to construct a new structure tensor. The latter is incorporated in a strain-energy function that accommodates both the mechanical and structural features of the material, extending our rotationally symmetric dispersion model (Gasser et al. 2006 J. R. Soc. Interface 3, 15-35. (doi:10.1098/rsif.2005.0073)). We provide specific ranges for the dispersion parameters and show how previous models can be deduced as special cases. We also provide explicit expressions for the stress and elasticity tensors in the Lagrangian description that are needed for a finite-element implementation. Material and structural parameters were obtained by fitting predictions of the model to experimental data obtained from human abdominal aortic adventitia. In a finite-element example, we analyse the influence of the fibre dispersion on the homogeneous biaxial mechanical response of aortic strips, and in a final example the non-homogeneous stress distribution is obtained for circumferential and axial strips under fixed extension. It has recently become apparent that this more general model is needed for describing the mechanical behaviour of a variety of fibrous tissues.

Quantitative assessment of collagen fibre orientations from two-dimensional images of soft biological tissues.

In this work, we outline an automated method for the extraction and quantification of material parameters characterizing collagen fibre orientations from two-dimensional images. Morphological collagen data among different length scales were obtained by combining the established methods of Fourier power spectrum analysis, wedge filtering and progressive regions of interest splitting. Our proposed method yields data from which we can determine parameters for computational modelling of soft biological tissues using fibre-reinforced constitutive models and gauge the length scales most appropriate for obtaining a physically meaningful measure of fibre orientations, which is representative of the true tissue morphology of the two-dimensional image. Specifically, we focus on three parameters quantifying different aspects of the collagen morphology: first, using maximum-likelihood estimation, we extract location parameters that accurately determine the angle of the principal directions of the fibre reinforcement (i.e. the preferred fibre directions); second, using a dispersion model, we obtain dispersion parameters quantifying the collagen fibre dispersion about these principal directions; third, we calculate the weighted error entropy as a measure of changes in the entire fibre distributions at different length scales, as opposed to their average behaviour. With fully automated imaging techniques (such as multiphoton microscopy) becoming increasingly popular (which often yield large numbers of images to analyse), our method provides an ideal tool for quickly extracting mechanically relevant tissue parameters which have implications for computational modelling (e.g. on the mesh density) and can also be used for the inhomogeneous modelling of tissues.