AI4COVID-19: AI enabled preliminary diagnosis for COVID-19 from cough samples via an app
Detection Principle
Others
Target
Cough
Testing Method Category
In_silico
Testing Method
Artificial intelligence screening solution
Testing Method - Additional Info
Deep Transfer Learning-based Multi Class classifier (DTL-MC); Classical Machine Learning-based Multi Class classifier (CML-MC); Deep Transfer Learning-based Binary Class classifier (DTL-BC)
Reported Performance
Accuracy: 92.64% (DTL-MC), 88.76% (CML-MC), 92.85% (DTL-BC); sensitivity: 89.14% (DTL-MC), 91.71 (CML-MC), 94.57 (DTL-BC); specificity: 96.67% (DTL-MC), 95.27 (CML-MC), 91.14% (DTL-BC); precision:89.91% (DTL-MC), 86.60% (CML-MC), 91.43 (DTL-BC)
Sample Size
543 cough samples (96 bronchitis, 130 pertussis, 70 COVID, 247 normal coughts)
Peer-reviewed
yes
The database contains available information from scientific literature that is being updated periodically. Please note that the provided information (as retrieved from analysed papers) is provided only for devices commercially available with CE-IVD mark. Acknowledgements