Classification of lung sounds using higher-order statistics: A divide-and-conquer approach

Naves, Raphaël; Barbosa, Bruno Henrique Groenner; Ferreira, Danton Diego

doi:10.1016/j.cmpb.2016.02.013

Cited by 72 publications

(34 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The study in [44] used higher order statistics to perform the classification of vesicular, fine and coarse crackle, and monophonic and polyphonic wheeze sounds. The classifier used was a combination of k-NN and NB.…”

Section: Resultsmentioning

confidence: 99%

Automatic adventitious respiratory sound analysis: A systematic review

2017

View full text Add to dashboard Cite

BackgroundAutomatic detection or classification of adventitious sounds is useful to assist physicians in diagnosing or monitoring diseases such as asthma, Chronic Obstructive Pulmonary Disease (COPD), and pneumonia. While computerised respiratory sound analysis, specifically for the detection or classification of adventitious sounds, has recently been the focus of an increasing number of studies, a standardised approach and comparison has not been well established.ObjectiveTo provide a review of existing algorithms for the detection or classification of adventitious respiratory sounds. This systematic review provides a complete summary of methods used in the literature to give a baseline for future works.Data sourcesA systematic review of English articles published between 1938 and 2016, searched using the Scopus (1938-2016) and IEEExplore (1984-2016) databases. Additional articles were further obtained by references listed in the articles found. Search terms included adventitious sound detection, adventitious sound classification, abnormal respiratory sound detection, abnormal respiratory sound classification, wheeze detection, wheeze classification, crackle detection, crackle classification, rhonchi detection, rhonchi classification, stridor detection, stridor classification, pleural rub detection, pleural rub classification, squawk detection, and squawk classification.Study selectionOnly articles were included that focused on adventitious sound detection or classification, based on respiratory sounds, with performance reported and sufficient information provided to be approximately repeated.Data extractionInvestigators extracted data about the adventitious sound type analysed, approach and level of analysis, instrumentation or data source, location of sensor, amount of data obtained, data management, features, methods, and performance achieved.Data synthesisA total of 77 reports from the literature were included in this review. 55 (71.43%) of the studies focused on wheeze, 40 (51.95%) on crackle, 9 (11.69%) on stridor, 9 (11.69%) on rhonchi, and 18 (23.38%) on other sounds such as pleural rub, squawk, as well as the pathology. Instrumentation used to collect data included microphones, stethoscopes, and accelerometers. Several references obtained data from online repositories or book audio CD companions. Detection or classification methods used varied from empirically determined thresholds to more complex machine learning techniques. Performance reported in the surveyed works were converted to accuracy measures for data synthesis.LimitationsDirect comparison of the performance of surveyed works cannot be performed as the input data used by each was different. A standard validation method has not been established, resulting in different works using different methods and performance measure definitions.ConclusionA review of the literature was performed to summarise different analysis approaches, features, and methods used for the analysis. The performance of recent studies showed a high agreement with conventiona...

show abstract

Section: Resultsmentioning

confidence: 99%

Automatic adventitious respiratory sound analysis: A systematic review

2017

View full text Add to dashboard Cite

show abstract

“…Os cinco melhores índices obtidos foram [12,13,16,23,157] ou [3,15,16,18, 155] reforçando a observação visual apresentada na figura 1, indicando que os primeiros índices do vetor da FFT são mais variados entre as classes e, portanto, seriam melhores critérios de separação entre as mesmas. Em [11] são apresentados diversos trabalhos de classificação onde dois deles, que utilizam metodologia similar, aliados ao uso do FFT como característica apresentam 80.9% e 95% como resultado correto de classificação.…”

Section: Resultsunclassified

“…Outro trabalho relevante é apresentado em [10] em que, por meio de redes neurais artificiais, classifica sons pulmonares em seis categorias: normal, sibilo, estertor, grasnidos, estridor ou ronco. Em [13] estatísticas de ordem superior são empregadas em uma abordagem dividir para conquistar para classificação de diversos tipos de sons pulmonares. Em [11] foi realizada uma revisão sistemática de uma gama de estudos em que é implementada análise computacional de sons pulmonares.…”

Section: Introductionunclassified

Detecção de Anormalidades em Sons Pulmonares Baseada em FFT e Máquinas de Vetores Suporte

Mota¹,

Dias²,

Ribeiro³

et al. 2018

Proceeding XIII Brazilian Congress on Computational Inteligence

View full text Add to dashboard Cite

“…Com a expectativa de atenuar o erro e dependência humana, durante o procedimento de ausculta e diagnósticos, a utilização de técnicas de reconhecimento de padrões tem sido empregada para superar estas lacunas (Naves, R., 2016;Dias et al, 2018;Mota et al, 2017). Por exemplo, em Amaral et al (2015), foi desenvolvido um sistema para tornar mais confiável a técnica de oscilação forçada (do inglês forced oscillation technique -FOT) na categorizacão do nível de obstrução das vias aéreas em pacientes com DPOC, basedo em três classificadores: o k-Nearest Neighbours (k-NN), Floresta Aleatória (RF) e Máquinas de Vetores de Suporte (SVM).…”

Section: Introductionunclassified

Classificação de Sons Pulmonares com Redes Convolucionais e SVM

Dias

Mota

Ribeiro

et al. 2019

Anais Do 14º Simpósio Brasileiro De Automação Inteligente

View full text Add to dashboard Cite

Pulmonary auscultation has a great importance for clinical evaluation, and requires adequate environment and an experienced evaluator. This study aims to present an acquisition system to classify three classes of lung sounds (vesicular, wheezy and crackles) using two techniques. The first is based on Fast Fourier Transform (FFT), Genetic Algorithms and Support Vector Machine (SVM). It was possible to classify lung sounds with sensitivity of 87.5% and specificity of 93.7%. The second technique uses the spectrograms with Convolutional Neural Networks (CNN) and has sensibility of 95.8% and the specificity of 87.5%, and is able to classify wheezes and crackles with 96% and 92 % of accuracy, respectively. Therefore, the SVM technique is superior in tracking and classifying lung sounds, at the cost of a higher rate of positive failures. In this wise, the digital stethoscope presented in this paper can be used to analyse lung sounds. Resumo: A ausculta pulmonar tem grande importância para a decisão diagnóstica na prática médica e requer ambiente apropriado e um avaliador experiente. Este trabalho tem por objetivo apresentar um sistema de aquisição e classificação de três classes de sons pulmonares que são: vesiculares, sibilos e estertores, utilizando duas técnicas diferentes. Na primeira técnica, Algoritmos Genéticos são utilizados para selecionar as melhores componentes da Transformada Rápida de Fourier (FFT) dos sinais em conjunto com Máquinas de Vetores Suporte (SVM). Foi possível a classificação dos sons pulmonares com sensibilidade de 87, 5% e especificidade de 93, 7%. A segunda técnica utiliza os espectrogramas dos sons e Redes Neurais Convolucionais (RNCs), nesse caso a sensibilidade foi de 95, 8% e especificidade de 87, 5%, eé capaz de classificar os sibilos com 96% e estertores com 92% de acertos. Logo a RNCé superior em detectar e classificar os sons anormais,às custas de um maioríndice de falsos positivos. Dessa forma, o estetoscópio digital apresentado pode ser empregado na análise de sons pulmonares.

show abstract

Classification of lung sounds using higher-order statistics: A divide-and-conquer approach

Cited by 72 publications

References 10 publications

Automatic adventitious respiratory sound analysis: A systematic review

Automatic adventitious respiratory sound analysis: A systematic review

Detecção de Anormalidades em Sons Pulmonares Baseada em FFT e Máquinas de Vetores Suporte

Classificação de Sons Pulmonares com Redes Convolucionais e SVM

Contact Info

Product

Resources

About