Abnormal Sounds Research Articles

Classification of respiratory sounds using improved convolutional recurrent neural network

Currently, auscultation using a stethoscope is performed for the diagnosis of respiratory diseases. Auscultation is a simple and non-invasive diagnostic method; however, the diagnostic results depend on the experience of the doctor, thereby rendering quantitative diagnosis difficult. Therefore, we herein propose a new automatic classification method based on deep learning algorithms for respiratory sounds to support the diagnosis of respiratory diseases. The proposed method comprises two stages. First, a spectrogram is generated by applying a short-time Fourier transform to the respiratory sound data. Subsequently, the obtained spectrogram is classified into normal and abnormal (three classes: crackle, wheeze, and both) respiratory sounds using an improved convolutional recurrent neural network. By classifying the respiratory sounds using the proposed method, the following results are obtained: sensitivity, 0.63; specificity,0.83; average score, 0.73; harmonic score, 0.72. Furthermore, the proposed method yields better accuracy compared with other methods.

A Retrospective Study on Bacteriology, Clinicopathologic and Radiographic Features in 28 Cats Diagnosed with Pyothorax.

Simple SummaryThe cat is a popular pet in Malaysia. Pyothorax is one of the most common respiratory conditions of cats in Malaysia. The etiological agents, particularly bacteria, of local feline pyothorax cases have never been studied. Therefore, in the present study we aimed to identify the bacteria involved in feline pyothorax cases, detect the antimicrobial susceptibility of the bacteria involved and study the radiographic findings related to these cases. Clinical records of cats diagnosed with pyothorax between 2013 and 2020 were analysed. The cats were presented with respiratory signs. Blood results showed significant inflammatory patterns and radiographs revealed different findings for each cat. The common bacteria and the antimicrobial susceptibility of the bacteria were determined in this study. Reporting the most common bacteria involved will give clinicians useful information that will further help in the selection of appropriate treatment and help clinicians understand the pathogenesis of pyothorax in the Malaysian cat population.This retrospective study aimed to determine the etiological, clinicopathological, and radiographic features and outcome of feline pyothorax cases. Medical records from twenty-eight cats with pyothorax aged from 4 months to 10 years (median 10 months) diagnosed between 2013 and 2020 were reviewed. Dyspnoea (75.0%), abnormal lung sounds (75.0%) and open-mouth breathing (64.3%) were the predominant respiratory signs. Leucocytosis (61.5%), particularly monocytosis (68.0%), and hyperglobulinaemia (65.4%) were among the most prominent findings in blood analysis. Bilateral pleural effusion was found in 67.9% of the thoracic radiographs. A total of 47.4% of the cytological samples revealed the presence of bacteria, while all had positive bacterial growth. Pasteurella multocida, E. coli, Streptococcus spp., and Staphylococcus spp. were the predominant aerobic bacteria isolated from pleural effusion samples. A chest tube was placed in 64.3% of the cats and 66.7% of cats with chest tubes survived. In total, 46.4% of cats with pyothorax recovered. Amoxicillin–clavulanate was the antimicrobial of choice against aerobic bacteria found in this study and should be given in combination with antimicrobials that cover anaerobic bacteria. Chest tube placement is crucial for treatment success. Cytological results and bacterial culture may not be consistent; thus, bacterial culture should be performed for every case.

Clinical and Ultrasonographic Studies in Bovine Respiratory Disease (BRD) Affected Calves

Bovine Respiratory Diseases (BRD) are one of the most common and important disease problem affecting newborn calves. The aim of this study was to use clinical respiratory score and thoracic ultrasonography as a screening method for the diagnosis of BRD in newborn calves. This study was carried out on 59 newborn calves in the Al-salhia farm for investment in Alsharqia Governorate, Egypt. These calves were divided into two groups. The first group (G1) including 16 healthy control calves and the second group (G2) including 43 newborn calves suffered from respiratory disorders. The most prominent clinical finding in BRD-affected calves was fever, nasal and ocular discharge, coughing with abnormal tracheal and lung sounds. According to the ultrasonographic examination of the lung, the lower respiratory tract disorders classified lesion into lobar pneumonia 12 calves (35.3%), lobular pneumonia 16 calves (47.1%), mixed pneumonia 3 calves (8.82%) in addition to lung abscess 3 calves (8.82%). Clinical respiratory score and thoracic ultrasound score showed a highly significant increase in BRD affected calves, meanwhile, CRS showed a non-significant change in calves suffered from upper and lower respiratory tract disorders. TUS revealed a non-significant change in calves with lobar and mixed pneumonia and in lung abscess, in addition to a significant increase in calves with lobar pneumonia when compared with lobular pneumonia. BRD affected calves had CRS ≥ five meanwhile calves with TUS ≥ two were considered sick.

Heart cardiac’s sounds signals segmentation by using the discrete wavelet transform (DWT)

The presence of abnormal sounds in one cardiac cycle, provide valuable information on various diseases.Early detection of various diseases is necessary; it is done by a simple technique known as: phonocardiography. The phonocardiography, based on registration of vibrations or oscillations of different frequencies, audible or not, that correspond to normal and abnormal heart sounds. It provides the clinician with a complementary tool to record the heart sounds heard during auscultation. The advancement of intracardiac phonocardiography, combined with signal processing techniques, has strongly renewed researchers’ interest in studying heart sounds and murmurs. This paper presents an algorithm based on the denoising by wavelet transform (DWT) and the Shannon energy of the PCG signal, for the detection of heart sounds (the first and second sounds, S1 and S2) and heart murmurs. This algorithm makes it possible to isolate individual sounds (S1 or S2) and murmurs to give an assessment of their average duration.

Uncemented total hip arthroplasty using alumina ceramic bearings at a minimum of 14years of follow-up.

The effectiveness of modern ceramic bearings has been well established in reducing the osteolysis associated with wear of the bearing surfaces in total hip arthroplasty (THA). However, there are limited mid- to long-term follow-up data for complications associated with ceramic bearings. This case series analyzed 124 consecutive primary uncemented THAs in 108 patients with a mean age of 61years using alumina ceramic-on-alumina ceramic bearing couples. Seventy THAs (56%) were evaluated at a minimum 14years of follow-up; the mean follow-up period was 16 ± 1years (14-20years). Kaplan-Meier survivorship was determined with revision surgery for any reason as the end point. Complications were recorded focusing on osteolysis, ceramic fracture, and abnormal sounds until the final follow-up. Clinical data were scored according to the Merle d'Aubigne and Postel hip score at 14years after THA. The survivorship was 93.5% (95% CI 86.7-97.0%) at 14-years postoperatively. Five patients (4.0%) underwent revision surgery due to instability or infection before 1-year postoperatively. Two patients (1.6%) underwent revision surgery due to ceramic liner fracture at 9- and 12-years postoperatively, respectively. There was no radiographic evidence of osteolysis. A total of 27 complications occurred: dislocations (n = 7/124), squeaking sounds (n = 3/124), clicking sounds (n = 6/124), ceramic liner fractures (n = 2/124), periprosthetic fractures (n = 4/124), deep infections (n = 2/124), transient sciatic nerve palsy (n = 2/124), and femoral stem breakage (n = 1/124). The Merle d'Aubigne and Postel hip score was 16.8 ± 1.4 points. The survivorship analysis demonstrates the uncemented THA using alumina ceramic bearings may provide favorable clinical outcome and can offer minimal wear at a minimum 14-year follow-up. Revision surgery was mostly required due to instability and infection in the short-term, and implant breakage in the mid- to long-term.

Intracoelomic Teratoma in an Eclectus Parrot (Eclectus roratus).

A 26-year-old, male eclectus parrot (Eclectus roratus) was presented to its primary care veterinarian for a 10-day history of increased respiratory effort, lethargy, and a single episode of regurgitation. Hepatomegaly, proventricular enlargement, cranial displacement of the ventriculus, and coelomic effusion were suspected based on a 1-view radiographic image, and the patient was referred for further evaluation. On presentation to the referral veterinary hospital, a distended coelom and dyspnea with abnormal respiratory sounds were abnormalities noted upon physical examination of the patient. The bird was hospitalized for supportive care, and diagnostic tests were performed. Initial radiographic images at the referral hospital revealed a large intracoelomic mass. A computed tomographic scan was performed; however, the origin of the mass could not be determined. A fine-needle aspirate and cytologic evaluation of the intracoelomic mass revealed a neoplastic process but no specific tissue type. Two days after presentation to the referral hospital, an exploratory coeliotomy to surgically resect the mass was attempted. The mass occupied most of the coelomic cavity, with multiple adhesions to internal organs. The mass was successfully resected; however, the patient destabilized and died despite resuscitation efforts. Histopathologic examination of submitted tissue from the mass with immunohistochemistry revealed mixed populations of neoplastic cells differentiated from 3 primordial germinal layers, confirming the diagnosis of teratoma. Teratomas appear to be a rare tumor in avian species but should be included in a list of differential disease diagnoses for abnormal tissue masses of unknown origin. Only 2 cases of teratomas have, to our knowledge, been reported in psittacine species.

Risk Factors for Tracheobronchomalacia in Preterm Infants With Bronchopulmonary Dysplasia.

Aim: To identify the risk factors associated with the development of tracheobronchomalacia (TBM) in preterm infants with bronchopulmonary dysplasia (BPD).Methods: This was a retrospective cohort study using chart reviews of preterm infants born at ≤ 36 week's gestation who underwent flexible fiberoptic bronchoscopy in a tertiary pediatric referral center between January 2015 and January 2020. Indications for the bronchoscopy examination included lobar atelectasis on plain chest film, persistent CO2 retention, recurrent extubation failure, or abnormal breathing sounds such as wheeze or stridor. Optimal cutoff values for each risk factor were also determined.Results: Fifty-eight preterm infants with BPD were enrolled, of whom 29 (50%) had TBM. There were no significant differences in gestational age and birth weight between those with and without TBM. Significantly more of the patients with TBM had severe BPD compared to those without TBM (68.9 vs. 20.6%, p < 0.001). Clinical parameters that were significantly different between the two groups were included in multivariate analysis. Among these factors, severe BPD was the most powerful risk factor for the development of TBM (odds ratio 5.57, 95% confidence interval 1.32–23.5, p = 0.019). The areas under the receiver operating characteristic curves for peak inspiratory pressure (PIP) and the duration of intubation were 0.788 and 0.75, respectively. The best predictive cutoff values of PIP and duration of intubation for TBM were 18.5 mmHg and 82 days, respectively.Conclusion: Preterm infants with severe BPD are at high risk for the development of TBM, and the risk is even higher in those who receive a higher PIP or are intubated for longer. Bronchoscopy examinations should be considered for the early diagnosis and management of TBM in infants with these risk factors.

Stress Cardiomyopathy Due to Exogenous Thyrotoxicosis From T3 Supplementation

Background: Wilson’s Temperature Syndrome (WTS) refers to a constellation of nonspecific symptoms, some of which include a low-normal body temperature, headaches, and fatigue. WTS was rejected by the American Thyroid Association as a valid medical diagnosis in 2005. Originally proposed in 1990 by Dr. E Denis Wilson, the etiology was theorized to be an impaired ability to convert T4 to T3, although this was not corroborated in thyroid lab abnormalities. Despite being publicly rebuffed as a true medical diagnosis, WTS has continued to gain traction amongst certain alternative medical groups and is treated with T3 supplementation. The development of a stress cardiomyopathy due to T3 supplementation is rare. Extrapolating from endogenous thyroid mediated stress cardiomyopathy and T4 supplementation induced stress cardiomyopathy, the pathophysiologic mechanism is likely excessive sympathetic activation. Clinical Case: A 58-year old female with no prior cardiac history presented to the emergency department with chest pain. Physical exam revealed a euvolemic appearing women, with intact and symmetric distal pulses, and a normal cardiac exam without murmurs or other abnormal heart sounds. There were no abnormal lung sounds. Her O2 sats were normal on room air. EKG and CXR were unremarkable. Her troponin was elevated (3.7 ng/mL, n< 0.034 ng/mL) and her BNP was elevated (4,568 pg/mL, n< 150 pg/mL). The patient was given aspirin and started on therapeutic heparin given concern for NSTEMI. Echocardiogram revealed an ejection fraction of 30% with hypokinesis of the entire apex and mid ventricle, raising concern for a stress cardiomyopathy.Coronary angiogram was performed which demonstrated no coronary artery disease. The patient underwent a cardiac MRI which confirmed a stress cardiomyopathy. Meanwhile, her laboratory workup was completed which revealed an undetectable TSH, a low T4 (0.53 ng/dL, n 0.7-1.8 ng/dL), and an elevated T3 (6.37 pg/mL, n 1.71-3.71 pg/mL). Patient endorsed taking oral liothyronine (T3) at doses of 5-40 mcg BID over the past 6 weeks per her alternative medicine provider for treatment of WTS. Her stress cardiomyopathy was presumed to be due to her exogenous thyrotoxicosis from T3 supplementation. She was counseled on the importance of cessation of T3 supplementation, and was started on heart failure medications. On hospital day 3, her T3 normalized to 2.73 pg/mL, and T4 remained low (0.4 pg/mL). Follow-up echocardiogram four months later demonstrated an ejection fraction that had improved to 45%. Conclusion: This case highlights the importance of physician awareness of alternative medicine diagnoses and treatment regimens that affect thyroid hormones and may cause harm to patients. This case is an important reminder of the effect thyroid hormones have on coronary vasculature, myocytes and myocardial function.

Food protein-induced enterocolitis syndrome presenting after necrotizing enterocolitis in a preterm neonate: a case report

When bloody stools occur in a very-low-birth-weight infant in the neonatal intensive care unit (NICU), necrotizing enterocolitis (NEC) is a prime consideration, though food protein-induced enterocolitis syndrome (FPIES) can be causative and is difficult to distinguish from NEC. Food allergy is an adverse reaction following exposure to food due to an abnormal immunologic response to food, and cow’s milk allergy (CMA) is the most likely form of food allergy in infants. The clinical features and proper management of patients with FPIES are important to differentiate FPIES from NEC. However, there are very few study reports of preterm infants presenting with food allergy-induced enterocolitis after NEC. Here, we report a case of a very-low-birth-weight infant born at 28 weeks of gestational age who developed recurrent episodes of bloody stools when he was fed cow’s milk or given breast milk fortified with milk after NEC recovery on day of life (DOL) 29, 46, and 54. A male preterm infant born at 28 weeks of gestational age presented with bloody stools on DOL 7. He was diagnosed with early-onset NEC with abdominal tenderness, sluggish bowel sounds, increased C-reactive protein (CRP) level and pneumatosis intestinalis (PI). After recovery from NEC on DOL 20, the infant developed three recurrent episodes of bloody stools after being fed cow’s milk or breast milk fortified with dairy milk. He was suspected of having recurrent episodes of NEC, but the infant was fairly healthy and did not present abdominal tenderness or abnormal bowel sounds on physical examination. Consecutive blood tests revealed normal CRP levels and increasing eosinophil levels. Abdominal radiograph revealed mild thickening of the small bowel, with no evidence of PI. The infant was finally diagnosed with FPIES in addition to NEC. After the infant received hydrolyzed formula, the bloody stool symptoms were finally resolved. Our case suggests that infants with recurrent episodes of bloody stools with increasing systemic eosinophils count should be considered for the diagnosis of FPIES with cow’s milk formula. Rapid improvement and non-progression of systemic symptoms and signs after removing exposure to milk protein may differentiate FPIES from NEC.

이상 호흡음 분류를 위한 계층적 어텐션 네트워크 모델

Recently the diagnosis of respiratory diseases based on respiratory-sound data has drawn much attention. However, no quantitative assessment method exists to detect abnormal respiratory sounds when making a diagnosis. Although existing studies have attempted to support doctors by providing them with machine learning results, there were limitations on explaining causal symptoms and establishing quantitative assessment. Therefore, a reliable method that can diagnose and explain the causal symptoms based on respiratory sound data is required. In this study, we propose using a hierarchical attention network for detecting abnormal respiratory sounds. The hierarchical attention network reflects hierarchical patterns of respiratory sounds and allows us to interpret the important feature of respiratory sounds. The experimental results showed that hierarchical attention network performed better than other existing methods in terms of sensitivity (ability of correctly detecting abnormal respiratory sounds) and explainability. We believe that the framework presented in this study can not only serve as a “second opinion” that can help doctors diagnose existing respiratory diseases, but also help doctors’ future research on unidentified respiratory diseases.

A low-cost IoT-based auscultation training device

&lt;span&gt;Auscultation training devices are needed by teachers and students in health schools to practice auscultation techniques. In this paper, a low-cost IoT-based auscultation training device has been developed using NodeMCU, four proximity sensors, metal as a stethoscope, a switch, an android smartphone, an earphone, and a phantom doll. The message queuing telemetry transport (MQTT) protocol has been used for data communication between NodeMCU and smartphones, therefore an auscultation training hardware can be used by many students who have auscultation training application on their smartphones that subscribe to topics. The results showed that an auscultation training device was able to detect a stethoscope. Auscultation training application on a smartphone successfully plays normal and abnormal breathing sounds based on subscribed topics. With a production cost of less than 15 USD, we offer an inexpensive IoT-based auscultation training device. &lt;/span&gt;

A case of hypercapnic respiratory failure.

A systematic work-up is important in case of hypercapnia. Pay attention to the shape of the flow–volume curve and any abnormal breathing sounds. In case of stridor, vocal cord paralysis should be suspected and, if confirmed, neurological investigations are advised. https://bit.ly/34APMi8

추진축과 스트러트 베어링 간 스틱슬립 비선형 마찰진동으로 인해 발생한 이상소음의 분석

추진축과 스트러트 베어링 간 스틱슬립 비선형 마찰진동으로 인해 발생한 이상소음의 분석

Evaluation of Overground Endoscopy Findings in Colombian Criollo Paso Horses

The dynamic upper airway functional (URT) problems of Colombian paso horses (CPH) have not been investigated in literature up to date. These horses perform special walking gaits with high poll flexion of the neck. Our goal is to evaluate the upper airway mechanics in CPH, showing abnormal respiratory sounds and poor performance during exercise. Resting and overground endoscopy was performed in 40 CPHs. Statistical analyses were performed using the sciPy package. One-tailed Fisher exact tests were used to check for positive contingency between each pair of upper respiratory tract disorder (P < .05). Arytenoid cartilage collapse was observed in 35 of 40 cases during exercise. Among these, dynamic laryngeal collapse (DLC) was the most significant finding, but ventromedial luxation of the apex of the corniculate process of the arytenoid and recurrent laryngeal neuropathy was also observed. Dorsal displacement of the soft palate was only detected in 4 of 40 cases. DLC was significantly associated with vocal cord collapse, nasopharyngeal collapse, and medial collapse of the margins of the epiglottis, and medial collapse of the margins of the epiglottis was associated with nasopharyngeal collapse. DLC is only a typical feature in some special breeds worldwide. We suggest that the extensive poll flexion and the relatively small laryngeal lumen and high intensitivity workload are the most important predisposing factors of DLC. Ventromedial luxation of the apex of the corniculate process of the arytenoid was over-represented in our caseload, compared with other studies. Whereas, despite the intensive workload, dorsal displacement of the soft palate was relatively uncommon. We presume that this could also be tailored to the high poll flexion performed during the special gaits.

Abnormal sound detection of washing machines based on convolution neural network in production environment

Convolutional Neural Networks (CNN) have been shown to have great advantages equally in the field of image and audio. In the field of abnormal sounds detection of household appliances in the production environment, the fundamental difficulty is to extract and recognise the features that can represent abnormal sounds effectively. However, due to the lack of knowledge reserve and the wide variety of data volume, appropriate feature extraction is not easy in the actual production process of home appliance products. In this paper, an end-to-end CNN deep model framework is designed by us for washing machine type rotating machinery data analysis, which can perform adaptive mining on the features existing in the original rotating mechanical data even under the influence of different rotational speeds and considerable noise. By validating on real data sets with different characteristics, the results show that the method can realise online fast training learning and offline testing. The test time is shorter than one second, and the highest test classification accuracy is 99.3%.

Abnormal sound detection of washing machines based on convolution neural network in production environment

Convolutional Neural Networks (CNN) have been shown to have great advantages equally in the field of image and audio. In the field of abnormal sounds detection of household appliances in the production environment, the fundamental difficulty is to extract and recognise the features that can represent abnormal sounds effectively. However, due to the lack of knowledge reserve and the wide variety of data volume, appropriate feature extraction is not easy in the actual production process of home appliance products. In this paper, an end-to-end CNN deep model framework is designed by us for washing machine type rotating machinery data analysis, which can perform adaptive mining on the features existing in the original rotating mechanical data even under the influence of different rotational speeds and considerable noise. By validating on real data sets with different characteristics, the results show that the method can realise online fast training learning and offline testing. The test time is shorter than one second, and the highest test classification accuracy is 99.3%.

One Class SVM を用いたベアリング異常音検知システム

In order for a machine to operate safely, it is one of the very important factors to quickly and accurately detect an abnormality. A skilled operator can make a diagnosis from the sound produced by the machine. However, constant monitoring by the operator is costly and inefficient. If abnormalities can be detected and maintained at an early stage, it will lead to avoiding serious failures, and cost reduction can be expected by long-term use of the machine. In this study, we proposed a system that uses machine learning to detect abnormalities from the operating sounds of machines. Instead of the conventional judgment method that requires abnormal sounds in advance, a system that judges them only from learning normal sound data is proposed. In order to confirm the effectiveness of the proposed system, we build a machine operation sound generator and conducted experiments by artificially generating abnormal noise of a bearing. We extracted features called Mel-Frequency Cepstrum Coefficient from the sounds, abnormal sounds were detected by using One Class Support Vector Machine. As a result, it was confirmed that the generated abnormal noise was detected, and it was found that the proposed system was effective for abnormality detection.

Using Deep Autoencoders for In-vehicle Audio Anomaly Detection

Current developments on self-driving cars have increased the interest on autonomous shared taxicabs. While most self-driving technologies focus on the outside environment, there is also a need to provide in-vehicle intelligence (e.g., detect health and safety issues related with the car occupants). Set within an R&D project focused on in-vehicle cockpit intelligence, the research presented in this paper addresses an unsupervised Acoustic Anomaly Detection (AAD) task. Since data is nonexistent in this domain, we first design an in-vehicle sound event data simulator that can realistically mix background audios (recorded from car driving trips) with normal (e.g., people talking, radio on) and abnormal (e.g., people arguing, cough) event sounds, allowing the generation of three synthetic in-vehicle sound datasets. Then, we explore two main sound feature extraction methods (based on a combination of three audio features and mel frequency energy coefficients) and propose a novel Long Short-Term Memory Autoencoder (LSTM-AE) deep learning architecture for in-vehicle sound anomaly detection. Competitive results were achieved by the proposed LSTM-AE when compared with two state-of-the-art methods, namely a dense Autoencoder (AE) and a two-stage clustering.

Abnormal Pulmonary Sounds Classification Algorithm using Convolutional Networks

In the world and in Peru, Acute Respiratory Infections are the main cause of death, especially in the most vulnerable population, children under 5 years of age and older adults. Pneumonia is the leading cause of death of children in the world. 60.2% of pneumonia cases affect children under 5 years of age. Thus, prevention and timely treatment of lung diseases are crucial to reduce infant mortality in Peru. Among the main problems associated with this high is percentage the lack of medical professionals and resources, especially in remote areas, such as Puno, Huancavelica and Arequipa, which experience temperatures as low as -20°C during the cold season. This study develops an algorithm based on computational neural networks to differentiate between normal and abnormal lung sounds. The initial base of 917 sounds was used, through a process of data augmentation, this base was increased to 8253 sounds in total, and this process was carried out due to the need of a large number of data for the use of computational neural networks. From each signal, features were extracted using three methods: MFCC, Melspectogram and STFT. Three models were generated, the first one to classify normal and abnormal, which obtained a training Accuracy of 1 and a testing accuracy of 0.998. The second one classifies normal sound, pneumonia and other abnormalities and obtained training Accuracy values of 0.9959 and a testing accuracy of 0.9885. Finally, we classified by specific ailment where we obtained a training Accuracy of 0.9967 and a testing accuracy of 0.9909. This research provides interesting findings about the diagnosis and classification of lung sounds automatically using convolutional neural networks, which is the beginning for the development of a platform to assess the risk of pneumonia in the first moment, thus allowing rapid care and referral that seeks to reduce mortality associated mainly with pneumonia.

Tourette disorder: A case report.

Tourette disorder: A case report.