Machine learning approach for automatic lungs sound diagnosis from pulmonary signals
Identifikátory výsledku
Kód výsledku v IS VaVaI
<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216305%3A26220%2F21%3APU141505" target="_blank" >RIV/00216305:26220/21:PU141505 - isvavai.cz</a>
Výsledek na webu
<a href="http://dx.doi.org/10.1109/TSP52935.2021.9522663" target="_blank" >http://dx.doi.org/10.1109/TSP52935.2021.9522663</a>
DOI - Digital Object Identifier
<a href="http://dx.doi.org/10.1109/TSP52935.2021.9522663" target="_blank" >10.1109/TSP52935.2021.9522663</a>
Alternativní jazyky
Jazyk výsledku
angličtina
Název v původním jazyce
Machine learning approach for automatic lungs sound diagnosis from pulmonary signals
Popis výsledku v původním jazyce
Chronic Respiratory Diseases (CRDs) are the most common diseases that affect people in today’s world. In COVID 19 pandemic many people are suffering from different types of respiratory diseases. There is a shortage of medical professionals and hence there is a requirement of artificial intelligence-based tools for automatic diagnosis of pulmonary diseases in the lungs. This paper presents a machine learning-based automatic classification method for the diagnosis of multiple pulmonary diseases from lung sounds. This work uses comprehensive lung sound categories labeled by a medical professional for use in machine learning-based classification. The proposed work uses four machine-learning classifiers (SVM, KNN, Naïve Bayes, and ANN) for the different discriminant features of lung sounds such as wheezing sound that can be used for diagnosis of asthma. For the detection of multiple lung sound in a noisy environment, data augmentation is used in training data and then trained the model where ANN using 5-fold cross-validation gives the average accuracy of 95.6%. The proposed method has low time complexity, is robust and non-invasive making it ideal for real-time applications to diagnose pulmonary diseases.
Název v anglickém jazyce
Machine learning approach for automatic lungs sound diagnosis from pulmonary signals
Popis výsledku anglicky
Chronic Respiratory Diseases (CRDs) are the most common diseases that affect people in today’s world. In COVID 19 pandemic many people are suffering from different types of respiratory diseases. There is a shortage of medical professionals and hence there is a requirement of artificial intelligence-based tools for automatic diagnosis of pulmonary diseases in the lungs. This paper presents a machine learning-based automatic classification method for the diagnosis of multiple pulmonary diseases from lung sounds. This work uses comprehensive lung sound categories labeled by a medical professional for use in machine learning-based classification. The proposed work uses four machine-learning classifiers (SVM, KNN, Naïve Bayes, and ANN) for the different discriminant features of lung sounds such as wheezing sound that can be used for diagnosis of asthma. For the detection of multiple lung sound in a noisy environment, data augmentation is used in training data and then trained the model where ANN using 5-fold cross-validation gives the average accuracy of 95.6%. The proposed method has low time complexity, is robust and non-invasive making it ideal for real-time applications to diagnose pulmonary diseases.
Klasifikace
Druh
D - Stať ve sborníku
CEP obor
—
OECD FORD obor
20203 - Telecommunications
Návaznosti výsledku
Projekt
—
Návaznosti
S - Specificky vyzkum na vysokych skolach
Ostatní
Rok uplatnění
2021
Kód důvěrnosti údajů
S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů
Údaje specifické pro druh výsledku
Název statě ve sborníku
44th International Conference on Telecommunications and Signal Processing (TSP)
ISBN
978-1-6654-2933-7
ISSN
—
e-ISSN
—
Počet stran výsledku
6
Strana od-do
366-371
Název nakladatele
IEEE
Místo vydání
Virtual Conference
Místo konání akce
Brno
Datum konání akce
26. 7. 2021
Typ akce podle státní příslušnosti
WRD - Celosvětová akce
Kód UT WoS článku
000701604600078