Table 5 Diagnostic accuracy of the deep learning model using all MostGraph-measured items as inputs to the model (by sex) and using only 10 measurement items as inputs (including age).

From: Pattern recognition of forced oscillation technique measurement results using deep learning can identify asthmatic patients more accurately than setting reference ranges

 

All MostGraph-measured items

10 inhaled and exhaled items

Diagnostic accuracy for healthy controls

Diagnostic accuracy for patients with asthma

Diagnostic accuracy for healthy controls

Diagnostic accuracy for patients with asthma

Both sexes

0.52 ± 0.08

0.84 ± 0.04

0.53 ± 0.12

0.85 ± 0.09

Male participants

0.48 ± 0.06

0.76 ± 0.15

0.36 ± 0.18

0.76 ± 0.14

Female participants

0.51 ± 0.11

0.84 ± 0.08

0.52 ± 0.13

0.85 ± 0.13

Back to article page