Table 3.

Summary table of recent studies found in literature for the use of machine/deep learning approaches in lung sounds classification

Study	No. patients	No. recordings	No. classes	Extracted features	Models	Performance
Aykanat et al. (2017)	1630	15,328	3: Normal. rale, rhonchus	MFCC/spectrograms	SVM/CNN	Accuracy: 80.00$\%$/80.00$\%$ Sensitivity: 89.00$\%$/79.00$\%$ Specificity: N/A
Bardou et al. (2018)	15	2141	7: Normal, monophonic wheeze polyphonic wheeze, stridor squawk, fine crackle, coarse crackle	Spectrograms	CNN	Accuracy: 95.56$\%$ Sensitivity: N/A Specificity: N/A
Shi et al. (2019)	384	1152	3: Normal, asthma, pneumonia	Spectrograms	VGG-BDGRU	Accuracy: 87.41$\%$ Sensitivity: N/A Specificity: N/A
Demir et al. (2020)	126	6898	4: Normal, crackles, wheezes crackles+wheezes	Spectrograms	CNN	Accuracy: 71.15$\%$ Sensitivity: 61.00$\%$ Specificity: 86.00$\%$
García-Ordás et al. (2020)	126	920	6: Normal, asthma, pneumonia BRON, COPD, respiratory tract infection	Spectrograms	CNN	Accuracy: N/A Sensitivity: 98.81$\%$ Specificity: 98.61$\%$
This study	213	1,483	6: Normal, asthma, pneumonia BRON, COPD. heart failure	Spatial and temporal (CNN + BDLSTM)	CNN + BDLSTM	Accuracy: 99.62$\%$ Sensitivity: 98.43$\%$ Specificity: 99.69$\%$