Figures
- FIGURE 1
A map of the 14 sites (A–N) used for acquiring the acoustic data.
- FIGURE 2
Optional cloud connectivity for the VoqX. BT: Bluetooth; EMR: electronic medical record.
- FIGURE 3
Flowchart of the pre-processing and machine-learning algorithm. SVM: support vector machine.
- FIGURE 4
Two features (out of 164) that strongly depend on infrasound: Mel-Frequency Cepstrum Coefficient 1 (MFCC1) as a function of magnitude of breathing frequency for COVID-19, normal and non-COVID-19: a) with infrasound and b) no infrasound.
- FIGURE 5
Visual sound signatures of the VoqX recorded from: a) a healthy subject, b) a COVID-19 pneumonia patient and c) a patient with non-COVID-19 pneumonia. The colours represent the intensity of breathing sounds (dB full scale).
- FIGURE 6
Receiver operating characteristic curves after classification with and without infrasound: a) detection of “silent” COVID-19 pneumonia and b) detection of non-COVID-19 pneumonia.
Tables
- TABLE 1
Selected patient characteristics and physical findings by lung auscultation
Subjects Age (years) Height (cm) Weight (kg) Smoker Gender (F/M) Lung auscultation Crackles (%) Wheezes (%) Crackles and wheezes (%) Reduced intensity breathing sound (%) Vesicular breathing sounds (%) Healthy 141 42±19 168±11 73±21 52 58/83 16 0 0 8 76 Non-COVID-19 pneumonia 61 62±17* 166±9#,* 81±20 38#,* 33/28 59#,* 7#,* 3#,* 3 28#,* COVID-19 pneumonia 164 58±12* 170±11 78±28 32 77/87 19* 4* 0 2 75 Data are presented as n or mean±sd, unless otherwise stated. F: female; M: male. *: p<0.05 versus healthy; #: p<0.05 versus COVID-19 pneumonia.
- TABLE 2
Performance measures from 12 runs of random sets
Specificity (%) Sensitivity NPV (%) PPV Accuracy Non-COVID-19 pneumonia (%) COVID-19 pneumonia (%) Non-COVID-19 pneumonia (%) COVID-19 pneumonia (%) Test group# With IFS 93±4** 70±7 97±2** 90±4** 90±4* 94±3** 92±1** No IFS 80±7 69±11 93±4 85±5 84±6 85±3 85±2 Cross-validation¶ With IFS 87±2** 71±7* 93±2** 88±3* 78±6** 89±1** 87±2** No IFS 75±4 65±8 85±3 79±2 68±5 80±2 78±2 Auscultation by acoustic stethoscope (95% CI)+ 76 (68–83) 72 (59–82) 25 (17–32) 43 (37–50) 71 (62–79) 52 (44–66) VoqX data are presented as mean±sd. IFS: infrasound; NPV: negative predictive value; PPV: positive predictive value. #: not in the learning process; ¶: optimisation of the learning process; +: all data. *: p<0.05; **: p<0.01 versus no IFS.
Supplementary Material
Please note: supplementary material is not edited by the Editorial Office, and is uploaded as it has been supplied by the author.
Features for machine learning and their p-values for separating the three groups 00152-2022.supplement
