Digital Stethoscope Research Articles

The Effect of Mindfulness Meditation Based on Spiritual Care with Rose Aromatherapy on Stress and Blood Pressure in Type 2 Diabetes Mellitus (DM) Patients

Background: Uncontrolled blood sugar levels in patients with type 2 diabetes mellitus can trigger various complications, one of which is macroangiopathy, which is a complication of large blood vessels that affects changes in blood pressure. Stress is considered a psychological factor that can increase blood pressure. Stressful conditions in patients with diabetes mellitus can affect the control of blood sugar levels. Aim: The purpose of this study was to analyze the effect of Mindfulness Meditation based on spiritual care with rose aromatherapy on stress and blood pressure in Type 2 Diabetes Mellitus (DM) sufferers. Methods: The design of this study used Quasy Experimental design with pre test and post test with control group . The population in this study were all type 2 Diabetes Mellitus sufferers. Sampling was done by purposive sampling. The total number of respondents in this study was 102, divided into 51 intervention groups and 51 control groups. The independent variables in this study were mindfulness meditation based on care and rose aromatherapy and the dependent variables in this study were stress and blood pressure. The instrument used in this study was the DASS 42 questionnaire and measuring blood pressure with a digital spygmomanometer and stethoscope. data analysis using paired t test and independent t test. The hypothesis is accepted if the p value <0.05. Results: statistical test on stress levels obtained a p value of 0.000 in the intervention group and obtained a p value of 0.003 in the control group. For systolic blood pressure, a p value of 0.000 was obtained in the intervention group and and p value of 0.011 was obtained in the control group. For diastolic blood pressure, a p value of 0.000 was obtained in the intervention group and a p value of 0.042 was obtained in the control group. Conclusion: Mindfulness Meditation therapy based on Spiritual Care with Rose Aromatherapy can reduce stress and systolic and diastolic blood pressure. So this research is highly recommended for nurses in carrying out independent assignments, especially in providing nonpharmacological therapy.

Evaluation of a digital stethoscope for electrocardiographic recording in donkeys: Preliminary results

The digital stethoscope (DS) is a cost-effective single-lead digital stethoscope that allows simultaneous electrocardiographic (ECG) and phonocardiographic recordings on a smartphone. Despite its application in small animals and horses, there are currently no studies on its use in donkeys. The aim of this study was to evaluate the use of a new smartphone-based DS device in recording ECG tracings in donkeys. Standard base-apex lead ECG (sECG) and single-lead DS ECG (dECG) were simultaneously recorded for at least 30 s. Both sECG and dECG tracings were analysed by the same operator, recording heart rate, ECG waves and intervals, and the presence and duration of artefacts. Thirty-seven donkeys were included. The dECG tracings were interpretable in all the animals (100 %). The results showed perfect agreement between the sECG and dECG data for the classification of heart rhythm and P-wave polarity. Strong agreement was found in the evaluation of heart rate calculated manually and automatically by the smartphone app, QRS complex polarity, T wave polarity, and duration of the PR interval. However, no agreement was found in the evaluation of P wave duration, QRS complex duration and amplitude, and T wave duration and amplitude. In conclusion, although this is only a preliminary study, the DS was a valid, practical, and easy to use electrocardiographic tool for recording good-quality ECG tracings to assess the ECGs of donkeys in the field.

SonicGuard Sensor-A Multichannel Acoustic Sensor for Long-Term Monitoring of Abdominal Sounds Examined through a Qualification Study.

Auscultation is a fundamental diagnostic technique that provides valuable diagnostic information about different parts of the body. With the increasing prevalence of digital stethoscopes and telehealth applications, there is a growing trend towards digitizing the capture of bodily sounds, thereby enabling subsequent analysis using machine learning algorithms. This study introduces the SonicGuard sensor, which is a multichannel acoustic sensor designed for long-term recordings of bodily sounds. We conducted a series of qualification tests, with a specific focus on bowel sounds ranging from controlled experimental environments to phantom measurements and real patient recordings. These tests demonstrate the effectiveness of the proposed sensor setup. The results show that the SonicGuard sensor is comparable to commercially available digital stethoscopes, which are considered the gold standard in the field. This development opens up possibilities for collecting and analyzing bodily sound datasets using machine learning techniques in the future.

Artificial intelligence-based framework to identify the abnormalities in the COVID-19 disease and other common respiratory diseases from digital stethoscope data using deep CNN.

The utilization of lung sounds to diagnose lung diseases using respiratory sound features has significantly increased in the past few years. The Digital Stethoscope data has been examined extensively by medical researchers and technical scientists to diagnose the symptoms of respiratory diseases. Artificial intelligence-based approaches are applied in the real universe to distinguish respiratory disease signs from human pulmonary auscultation sounds. The Deep CNN model is implemented with combined multi-feature channels (Modified MFCC, Log Mel, and Soft Mel) to obtain the sound parameters from lung-based Digital Stethoscope data. The model analysis is observed with max-pooling and without max-pool operations using multi-feature channels on respiratory digital stethoscope data. In addition, COVID-19 sound data and enriched data, which are recently acquired data to enhance model performance using a combination of L2 regularization to overcome the risk of overfitting because of less respiratory sound data, are included in the work. The suggested DCNN with Max-Pooling on the improved dataset demonstrates cutting-edge performance employing a multi-feature channels spectrogram. The model has been developed with different convolutional filter sizes (, , , , and ) that helped to test the proposed neural network. According to the experimental findings, the suggested DCNN architecture with a max-pooling function performs better to identify respiratory disease symptoms than DCNN without max-pooling. In order to demonstrate the model's effectiveness in categorization, it is trained and tested with the DCNN model that extract several modalities of respiratory sound data.

An Accelerometer-Based Wearable Patch for Robust Respiratory Rate and Wheeze Detection Using Deep Learning.

Wheezing is a critical indicator of various respiratory conditions, including asthma and chronic obstructive pulmonary disease (COPD). Current diagnosis relies on subjective lung auscultation by physicians. Enabling this capability via a low-profile, objective wearable device for remote patient monitoring (RPM) could offer pre-emptive, accurate respiratory data to patients. With this goal as our aim, we used a low-profile accelerometer-based wearable system that utilizes deep learning to objectively detect wheezing along with respiration rate using a single sensor. The miniature patch consists of a sensitive wideband MEMS accelerometer and low-noise CMOS interface electronics on a small board, which was then placed on nine conventional lung auscultation sites on the patient's chest walls to capture the pulmonary-induced vibrations (PIVs). A deep learning model was developed and compared with a deterministic time-frequency method to objectively detect wheezing in the PIV signals using data captured from 52 diverse patients with respiratory diseases. The wearable accelerometer patch, paired with the deep learning model, demonstrated high fidelity in capturing and detecting respiratory wheezes and patterns across diverse and pertinent settings. It achieved accuracy, sensitivity, and specificity of 95%, 96%, and 93%, respectively, with an AUC of 0.99 on the test set-outperforming the deterministic time-frequency approach. Furthermore, the accelerometer patch outperforms the digital stethoscopes in sound analysis while offering immunity to ambient sounds, which not only enhances data quality and performance for computational wheeze detection by a significant margin but also provides a robust sensor solution that can quantify respiration patterns simultaneously.

Algorithm for predicting valvular heart disease from heart sounds in an unselected cohort.

This study aims to assess the ability of state-of-the-art machine learning algorithms to detect valvular heart disease (VHD) from digital heart sound recordings in a general population that includes asymptomatic cases and intermediate stages of disease progression. We trained a recurrent neural network to predict murmurs from heart sound audio using annotated recordings collected with digital stethoscopes from four auscultation positions in 2,124 participants from the Tromsø7 study. The predicted murmurs were used to predict VHD as determined by echocardiography. The presence of aortic stenosis (AS) was detected with a sensitivity of 90.9%, a specificity of 94.5%, and an area under the curve (AUC) of 0.979 (CI: 0.963-0.995). At least moderate AS was detected with an AUC of 0.993 (CI: 0.989-0.997). Moderate or greater aortic and mitral regurgitation (AR and MR) were predicted with AUC values of 0.634 (CI: 0.565-703) and 0.549 (CI: 0.506-0.593), respectively, which increased to 0.766 and 0.677 when clinical variables were added as predictors. The AUC for predicting symptomatic cases was higher for AR and MR, 0.756 and 0.711, respectively. Screening jointly for symptomatic regurgitation or presence of stenosis resulted in an AUC of 0.86, with 97.7% of AS cases (n = 44) and all 12 MS cases detected. The algorithm demonstrated excellent performance in detecting AS in a general cohort, surpassing observations from similar studies on selected cohorts. The detection of AR and MR based on HS audio was poor, but accuracy was considerably higher for symptomatic cases, and the inclusion of clinical variables improved the performance of the model significantly.

Evaluation of arteriovenous fistula for hemodialysis with a new generation digital stethoscope: a pilot study.

The management of complications of arteriovenous fistula (AVF) for hemodialysis, principally stenosis, remains a major challenge for clinicians with a substantial impact on health resources. Stenosis not infrequently preludes to thrombotic events with the loss of AVF functionality. A functioning AVF, when listened by a stethoscope, has a continuous systolic-diastolic low-frequency murmur, while with stenosis, the frequency of the murmur increases and the duration of diastolic component decreases, disappearing in severe stenosis. These evidences are strictly subjective and dependent from operator skill and experience. New generation digital stethoscopes are able to record sound and subsequently dedicated software allows to extract quantitative variables that characterize the sound in an absolutely objective and repeatable way. The aim of our study was to analyze with an appropriate software sounds from AVFs taken by a commercial digital stethoscope and to investigate the potentiality to develop an objective way to detect stenosis. Between September 2022 and January 2023, 64 chronic hemodialysis (HD) patients were screened by two blinded experienced examiners for recognized criteria for stenosis by Doppler ultrasound (DUS) and, consequently, the sound coming from the AVFs using a 3M™ Littmann® CORE Digital Stethoscope 8570 in standardized sites was recorded. The sound waves were transformed into quantitative variables (amplitude and frequency) using a sound analysis software. The practical usefulness of the core digital stethoscope for a quick identification of an AVF stenosis was further evaluated through a pragmatic trial. Eight young nephrologist trainees underwent a simple auscultatory training consisting of two sessions of sound auscultation focusing two times on a "normal" AVF sound by placing the digital stethoscope on a convenience site of a functional AVF. In 48 patients eligible, all sound components displayed, alone, a remarkable diagnostic capacity. More in detail, the AUC of the average power was 0.872 [95% CI 0.729-0.951], while that of the mean normalized frequency was 0.822 [95% 0.656-0.930]. From a total of 32 auscultations (eight different block sequences, each one comprising four auscultations), the young clinicians were able to identify the correct sound (stenosis/normal AVF) in 25 cases, corresponding to an overall accuracy of 78.12% (95% CI 60.03-90.72%). The analysis of sound waves by a digital stethoscope permitted us to distinguish between stenotic and no stenotic AVFs. The standardization of this technique and the introducing of data in a deep learning algorithm could allow an objective and fast method for a frequent monitoring of AVF.

Automatic detection of solar cell surface defects in electroluminescence images based on YOLOv8 algorithm

<p>Digital stethoscopes over recent years have gained acceptance among pulmonologists to perform auscultations due to their advantages over traditional stethoscopes. During the previous decade, researchers have prominently contributed to the development of algorithms aimed at enabling objective diagnosis of respiratory sounds and conditions, thereby affording individuals lacking medical expertise the capability to auscultate themselves. However, auscultation requires the per sonnel to be aware of the optimal chest position to place the device for a reliable diagnosis as well. This study aims to identify the optimal chest position to place a digital stethoscope’ s diaphragm to objectively diagnose chronic obstructive pulmonary d isease (COPD). Lung sound recordings from seven chest positions with data available in the ICBHI 2017 database namely, anterior left (Al), anterior right (Ar), lateral left (Ll), lateral right (Lr), posterior left (Pl), posterior right (Pr) and trachea (Tc), were analyzed in this study. COPD+ and COPD - at diagnosis, each chest position was done objectively using mel frequency cepstral coefficients (MFCC) features and machine learning models namely support vector machine (SVM) and decision tree (DT). The res ults indicate that the Pr chest position offers superior precision, recall, and F1 - score, with a recognition accuracy of 99.7% in COPD screening.</p>

Optimal chest position of auscultation for chronic obstructive pulmonary disease diagnosis using machine learning

Digital Stethoscopes over recent years have gained acceptance among pulmonologists to perform auscultations due to their advantages over traditional stethoscopes. During the previous decade, researchers have prominently contributed to the development of algorithms aimed at enabling objective diagnosis of respiratory sounds and conditions, thereby affording individuals lacking medical expertise the capability to auscultate themselves. However, auscultation requires the personnel to be aware of the optimal chest position to place the device for a reliable diagnosis as well. This study aims to identify the optimal chest position to place a digital stethoscope's diaphragm to objectively diagnose Chronic Obstructive Pulmonary Disease (COPD). Lung sound recordings from seven chest positions with data available in the ICBHI 2017 database namely, Anterior left (Al), Anterior right (Ar), Lateral left (Ll), Lateral right (Lr), Posterior left (Pl), Posterior right (Pr) and Trachea (Tc), were analyzed in this study. COPD+ and COPD- at diagnosis, each chest position was done objectively using Mel-Frequency Cepstral Coefficients (MFCC) features and machine learning models namely Support Vector Machine and Decision Tree. The results indicate that the Posterior right (Pr) chest position offers superior precision, recall, and F1 score, with a recognition accuracy of 99.7% in COPD screening.

Detection and management of latent rheumatic heart disease: a narrative review

Rheumatic heart disease (RHD) is a public health concern in many developing nations around the world. Early detection of latent or subclinical RHD can help in reversing mild lesions, retarding disease progression, reducing morbidity and mortality, and improving the quality of life of patients. Echocardiography is the gold-standard method for screening and confirming latent RHD cases. The rates and determinants of progression of latent RHD cases as assessed by echocardiography have been found to be variable through studies. Even though latent RHD has a slow rate of progression, the rate of progression of its subtype, ʻdefiniteʼ RHD, is substantial. A brief training of nonexpert operators on the use of handheld echocardiography with a simplified protocol is an important strategy to scale up the screening program to detect latent cases. Newer advancements in screening, such as deep-learning digital stethoscopes and telehealth services, have provided an opportunity to expand screening programs even in resource-constrained settings. Newer studies have established the efficacy and safety profile of secondary antibiotic prophylaxis in latent RHD. The concerned authorities in endemic regions of the world should work on improving the availability and accessibility of antibiotic prophylaxis.

Cardiac Auscultation Using Artificial Intelligence on Different Devices

Having to pay to go to a doctor's office and pay for a medical professional to use a stethoscope is costly and inconvenient. A mobile solution that is cheap and uses a medium that is widespread will make diagnoses more accessible. The objective of our research was to assess how feasible and accurate a mobile device solution to cardiac auscultation is, compared to a digital stethoscope. We used a convolutional neural network-based solution, which used the heart sound audio, collected with a digital stethoscope and smartphone, graphed out on a spectrogram for input. We trained two convolutional neural networks, one on the digital stethoscope audio and the other on the smartphone audio. To analyze the outputs, we used the metrics accuracy, recall, precision, and f1 score. We then compared the outputs of the model trained on the digital stethoscope audio versus the model trained on the smartphone audio. The model trained on the smartphone data typically performed 15% worse than the model trained on stethoscope data in terms of accuracy. Based off these results alone, the hardware technology in phones is still not advanced enough to reliably diagnose with machine learning.

Increasing detection of heart valve disease in Farnborough Primary Care Network using auscultation artificial intelligence in the community pharmacy setting

Abstract Introduction Cardiovascular disease causes a quarter of all deaths in the UK, making it a top NHS priority1. One such cardiovascular disease, heart valve disease (HVD), is serious with worse outcomes than many cancers2. Yet diagnosed early, before a patient deteriorates, HVD can be treated effectively with valve replacement or repair. A large proportion of patients however are estimated to be undiagnosed or untreated each year in England (≤53% with severe symptomatic aortic stenosis, ≤41% with moderate/severe mitral valve regurgitation, ≤81% with moderate or greater tricuspid valve regurgitation)3. HVD is often detected when GPs identify a heart murmur upon cardiac auscultation. Patients are referred to echocardiography to confirm diagnosis and severity. By using digital stethoscope to perform auscultation, community pharmacists have the potential to assist with and increase the detection of HVD. Aim To determine if community pharmacists can identify undiagnosed HVD using digital auscultation with eMurmur® stethoscope artificial intelligence (AI), and compare subsequent echocardiography referrals against those from GPs. Methods A community echocardiography service and pharmacy within a single PCN were identified based on interest, capacity, and infrastructure to undergo the service evaluation (SE). Pharmacy patients aged >75 years, or with type 2 diabetes, angina/myocardial infarction, atrial fibrillation or high blood pressure, were invited for HVD assessment by the pharmacist. Patients with known HVD or who had an echocardiogram in the past 10 years were excluded. The pharmacist carried out digital stethoscope auscultation and where systolic murmur was detected, patients were referred to a community clinic for full compensation echocardiography. Patient demographic, comorbidity, auscultation, and echocardiography results were collected. Clinic referrals from the community pharmacy were contrasted with referrals from GP practices to compare the number of appropriate referrals over the same period. Results Community pharmacy detected murmur in 39/86 (45%) patients, who were then referred to community echocardiography (mean age 77.8 years). Echocardiography identified: 8 (21%) moderate/severe HVD, 9 (23%) mild HVD, 22 (56%) ‘normal’ or ‘trivial’ HVD. During the SE, 24 referrals were received from GP practices (mean age 65 years): 2 (8%) moderate/severe HVD, 8 (33%) mild HVD, 14 (58%) ‘normal’ or ‘trivial’ HVD. Pharmacy referred 70% more patients with non-trivial HVD than GP practices (17 vs 10 patients). Discussion/Conclusion This SE demonstrates that community pharmacists can identify undiagnosed HVD, in this case referring more patients with non-trivial HVD to echocardiography than GPs. The rapid data sharing enabled by this technology supports quick decision-making for patients with critical disease requiring immediate treatment. Pharmacists can therefore play an important role as a gateway to accessing potentially lifesaving treatment. Patients appeared comfortable attending a digital auscultation service in a community pharmacy setting. Such a proactive service may help reduce the number of people with severe HVD who go undiagnosed and untreated. Limitations of this SE include the small sample size, rendering statistical analysis inappropriate. Additionally, the analysis indicated the number of GP referrals for echocardiography but not how many patients underwent auscultation overall. Therefore we were unable to compare efficiency of murmur detection.

Exploring classical machine learning for identification of pathological lung auscultations

The use of machine learning in biomedical research has surged in recent years thanks to advances in devices and artificial intelligence. Our aim is to expand this body of knowledge by applying machine learning to pulmonary auscultation signals. Despite improvements in digital stethoscopes and attempts to find synergy between them and artificial intelligence, solutions for their use in clinical settings remain scarce. Physicians continue to infer initial diagnoses with less sophisticated means, resulting in low accuracy, leading to suboptimal patient care. To arrive at a correct preliminary diagnosis, the auscultation diagnostics need to be of high accuracy. Due to the large number of auscultations performed, data availability opens up opportunities for more effective sound analysis. In this study, digital 6-channel auscultations of 45 patients were used in various machine learning scenarios, with the aim of distinguishing between normal and abnormal pulmonary sounds. Audio features (such as fundamental frequencies F0-4, loudness, HNR, DFA, as well as descriptive statistics of log energy, RMS and MFCC) were extracted using the Python library Surfboard. Windowing, feature aggregation, and concatenation strategies were used to prepare data for machine learning algorithms in unsupervised (fair-cut forest, outlier forest) and supervised (random forest, regularized logistic regression) settings. The evaluation was carried out using 9-fold stratified cross-validation repeated 30 times. Decision fusion by averaging the outputs for a subject was also tested and found to be helpful. Supervised models showed a consistent advantage over unsupervised ones, with random forest achieving a mean AUC ROC of 0.691 (accuracy 71.11%, Kappa 0.416, F1-score 0.675) in side-based detection and a mean AUC ROC of 0.721 (accuracy 68.89%, Kappa 0.371, F1-score 0.650) in patient-based detection.

Wireless broadband acousto-mechanical sensing system for continuous physiological monitoring.

The human body generates various forms of subtle, broadband acousto-mechanical signals that contain information on cardiorespiratory and gastrointestinal health with potential application for continuous physiological monitoring. Existing device options, ranging from digital stethoscopes to inertial measurement units, offer useful capabilities but have disadvantages such as restricted measurement locations that prevent continuous, longitudinal tracking and that constrain their use to controlled environments. Here we present a wireless, broadband acousto-mechanical sensing network that circumvents these limitations and provides information on processes including slow movements within the body, digestive activity, respiratory sounds and cardiac cycles, all with clinical grade accuracy and independent of artifacts from ambient sounds. This system can also perform spatiotemporal mapping of the dynamics of gastrointestinal processes and airflow into and out of the lungs. To demonstrate the capabilities of this system we used it to monitor constrained respiratory airflow and intestinal motility in neonates in the neonatal intensive care unit (n = 15), and to assess regional lung function in patients undergoing thoracic surgery (n = 55). This broadband acousto-mechanical sensing system holds the potential to help mitigate cardiorespiratory instability and manage disease progression in patients through continuous monitoring of physiological signals, in both the clinical and nonclinical setting.

Exploring Microphone Technologies for Digital Auscultation Devices.

The aim of this work is to present a preliminary study for the design of a digital auscultation system, i.e., a novel wearable device for patient chest auscultation and a digital stethoscope. The development and testing of the electronic stethoscope prototype is reported with an emphasis on the description and selection of sound transduction systems and analog electronic processing. The focus on various microphone technologies, such as micro-electro-mechanical systems (MEMSs), electret condensers, and piezoelectronic diaphragms, intends to emphasize the most suitable transducer for auscultation. In addition, we report on the design and development of a digital acquisition system for the human body for sound recording by using a modular device approach in order to fit the chosen analog and digital mics. Tests were performed on a designed phantom setup, and a qualitative comparison between the sounds recorded with the newly developed acquisition device and those recorded with two commercial digital stethoscopes is reported.

Implementing artificial intelligence-based cardio-obstetric screening in a diverse patient population: a prospective study

Abstract Background Cardiomyopathy is a leading cause of pregnancy-related mortality and represents the number one cause of death in the late postpartum period. Purpose The goal of this study was to prospectively evaluate the effectiveness of an artificial intelligence-based electrocardiogram (AI-ECG) and a point-of-care artificial intelligence-enhanced digital stethoscope for cardiomyopathy screening among pregnant and postpartum women. Methods We conducted a single-arm prospective study of pregnant and postpartum women (up to 12 months following delivery) enrolled at three sites between October 2021 and October 2022. Study participants completed a standard 12-lead ECG, ECG + phonocardiogram recordings with a digital stethoscope, a demographic questionnaire, and a transthoracic echocardiogram (TTE) within a 24-hour period. Diagnostic performance of the AI-ECG and digital stethoscope for cardiomyopathy detection, defined as left ventricular ejection fraction (LVEF) < 45%, was evaluated using the area under the curve (AUC) and standard measures of diagnostic performance. Results We included 100 consecutive pregnant and postpartum women who provided informed consent and had at least one ECG recorded, completed the questionnaire, and had an echocardiogram performed. All participants self-identified as female, and 41% were recruited from a federally qualified health center. The median age of participants was 31 years (Q1: 27, Q3: 34). Thirty-eight percent identified as non-Hispanic White, 32% as non-Hispanic Black, 21% as Hispanic or Latino, 6% as Asian, and 3% as Other (1 Native Hawaiian and 2 multiracial). Five percent (5/100) had LVEF <45%. The 12-lead AI-ECG model demonstrated close to perfect discrimination (accuracy 99%; AUC 1.00, sensitivity 100%) for detection of cardiomyopathy with 1 false positive screen (LVEF 49%), Figure 1A. The digital stethoscope (angled position) had an AUC of 0.99 (95% CI: 0.97, 1.00) for detection of LVEF <45%. Sensitivity, specificity, positive predictive value, and negative predictive value for the digital stethoscope were 60%, 98%, 60% and 98%, respectively for detection of LVEF <45% (Figure 1B & 1C). Conclusions In this small prospective study, we found that 5% of pregnant and postpartum women had cardiomyopathy, a potentially life-threatening condition, and that a rapid, inexpensive, ubiquitous, massively scalable point-of-care test - the ECG - showed promising results, identifying it with an AUC > 0.98 with the application of artificial intelligence.Figure 1

Detection of aortic stenosis using built-in microphones of commercially available smartphones

Abstract Background According to recent data, asymptomatic patients with severe aortic stenosis benefit from early treatment. Hence, widely applicable methods to identify patients with severe aortic stenosis in an asymptomatic state may be desirable in the future. While the diagnosis of aortic stenosis is currently based on echocardiography, population-wide use is not feasible due to resource constraints. With recent technological advances in built-in microphones of commercially available smartphones, digital auscultation may provide a viable alternative. Methods As part of this study, a smartphone application to perform digital auscultation was developed. To ensure high-quality recordings real-time signal analysis is implemented. The 8 kHz sampling rate of the signal encompasses all typical frequencies of cardiac auscultation while keeping resource usage minimal. To prevent frequency leakage the signal is attenuated using a Blackman-Harris window. After applying Fast Fourier Transformation to the windowed signal, frequencies around 100 Hz and 250 Hz are amplified to match the frequency response of traditional stethoscopes. The resulting signal amplitudes are then visualized. The application provides functionality for storing metadata about each patient in addition to the raw audio recordings. To verify the feasibility of using the described application to detect severe aortic stenosis, this study aims to enroll 50 persons with no valvular heart disease and 50 persons with severe aortic stenosis. Initial audio recordings by the smartphone application and of a digital stethoscope are compared by visualizing the frequency spectrum using the Mel Frequency Cepstral Coefficients. Results The peaks of the first and second heart sound at around 50 Hz are both clearly identifiable, as is the crescendo-decrescendo murmur typically associated with aortic stenosis in the frequency range of 100-600 Hz (Figure 1). This is true for both smartphone and stethoscope recordings, even though the built-in microphone of smartphones results in a lower signal-to-noise ratio. Nevertheless, the quality of the signal is robust enough to clearly indicate the presence of aortic stenoses. Conclusions The initial findings of this study indicate that a smartphone application using built-in microphones is a promising approach for identifying severe aortic stenoses. This approach has the potential to be a reliable and cost-effective screening tool for early detection. Further development of automated algorithms may enhance the tool's effectiveness and accessibility.Figure 1

Abstract 13244: Real World Evaluation of an Artificial Intelligence Enabled Digital Stethoscope for Detecting Undiagnosed Valvular Heart Disease in Primary Care

Background: Clinically significant valvular heart disease (VHD) affects 1 in 10 adults over 65. Cardiac auscultation, the current point of care clinical standard, has shown a 44% sensitivity for detecting VHD, leaving most patients undiagnosed. A deep learning based artificial intelligence (AI) platform that detects cardiac murmurs associated with clinically significant VHD (moderate severity or higher, confirmed by echocardiogram) was used to assess its real-time impact on the point of care. Hypothesis: Use of AI during point of care auscultation may lead to higher discovery rates and earlier diagnosis of VHD compared to conventional practice. Aim: Evaluate real world performance of AI vs. primary care physician (PCP) auscultation in identifying previously undiagnosed VHD. Methods: This blinded, prospective study enrolled 369 patients aged 50 and older without a prior VHD diagnosis or history of murmur from three different primary care clinics. Four-point cardiac auscultation was performed on each patient by (1) PCP with a standard stethoscope, and (2) coordinator with an Eko digital stethoscope to collect phonocardiogram (PCG) data for analysis by the AI. Each patient received an echocardiogram to confirm whether clinically significant VHD was present, and their PCG recordings were reviewed by an outside expert panel to confirm whether audible murmurs were present. PCP and expert panel were blinded to AI and echocardiogram results. Performance metrics were calculated for both PCP auscultation and the AI in detecting audible VHD. Results: The AI showed over twofold improvement in sensitivity vs. PCP auscultation for detecting audible VHD (94.1% vs. 41.2%) with limited impact on specificity (84.5% vs. 95.5%). The AI identified 22 patients with moderate-or-greater VHD who were previously undiagnosed, while PCPs identified eight previously undiagnosed patients with VHD. Conclusion: When applied in a primary care setting, a digital stethoscope with structural murmur detection AI showed meaningful impact on new discovery of VHD as compared to conventional practice. These results suggest that AI-powered auscultation at the point of care may significantly increase earlier VHD discovery, facilitate appropriate patient care, and improve outcomes.

Pulmonary disease detection and classification in patient respiratory audio files using long short-term memory neural networks.

In order to improve the diagnostic accuracy of respiratory illnesses, our research introduces a novel methodology to precisely diagnose a subset of lung diseases using patient respiratory audio recordings. These lung diseases include Chronic Obstructive Pulmonary Disease (COPD), Upper Respiratory Tract Infections (URTI), Bronchiectasis, Pneumonia, and Bronchiolitis. Our proposed methodology trains four deep learning algorithms on an input dataset consisting of 920 patient respiratory audio files. These audio files were recorded using digital stethoscopes and comprise the Respiratory Sound Database. The four deployed models are Convolutional Neural Networks (CNN), Long Short-Term Memory (LSTM), CNN ensembled with unidirectional LSTM (CNN-LSTM), and CNN ensembled with bidirectional LSTM (CNN-BLSTM). The aforementioned models are evaluated using metrics such as accuracy, precision, recall, and F1-score. The best performing algorithm, LSTM, has an overall accuracy of 98.82% and F1-score of 0.97. The LSTM algorithm's extremely high predictive accuracy can be attributed to its penchant for capturing sequential patterns in time series based audio data. In summary, this algorithm is able to ingest patient audio recordings and make precise lung disease predictions in real-time.

Deep Learning Algorithms to Detect Murmurs Associated With Structural Heart Disease.

Background The success of cardiac auscultation varies widely among medical professionals, which can lead to missed treatments for structural heart disease. Applying machine learning to cardiac auscultation could address this problem, but despite recent interest, few algorithms have been brought to clinical practice. We evaluated a novel suite of Food and Drug Administration-cleared algorithms trained via deep learning on >15 000 heart sound recordings. Methods and Results We validated the algorithms on a data set of 2375 recordings from 615 unique subjects. This data set was collected in real clinical environments using commercially available digital stethoscopes, annotated by board-certified cardiologists, and paired with echocardiograms as the gold standard. To model the algorithm in clinical practice, we compared its performance against 10 clinicians on a subset of the validation database. Our algorithm reliably detected structural murmurs with a sensitivity of 85.6% and specificity of 84.4%. When limiting the analysis to clearly audible murmurs in adults, performance improved to a sensitivity of 97.9% and specificity of 90.6%. The algorithm also reported timing within the cardiac cycle, differentiating between systolic and diastolic murmurs. Despite optimizing acoustics for the clinicians, the algorithm substantially outperformed the clinicians (average clinician accuracy, 77.9%; algorithm accuracy, 84.7%.) Conclusions The algorithms accurately identified murmurs associated with structural heart disease. Our results illustrate a marked contrast between the consistency of the algorithm and the substantial interobserver variability of clinicians. Our results suggest that adopting machine learning algorithms into clinical practice could improve the detection of structural heart disease to facilitate patient care.