Table 3

Test set precision, recall and F1 scores for each input variable combination when using data from the first 477 patients

Combination of input variablesLung patternPrecision for test set (%)Recall for test set (%)F1 score for test set (%)
Biometrics with spirometry and plethysmographyNormal91.60±2.9988.30±4.3589.87±3.10
Obstructive73.50±6.85100.0084.56±4.56
Restrictive100.0091.90±2.8595.76±1.49
Obstructive and restrictive100.0096.00±8.4397.78±4.68
Biometrics with spirometryNormal90.70±6.3492.60±4.9091.36±2.22
Obstructive89.60±11.6785.90±17.2186.28±10.87
Restrictive89.90±6.3788.00±7.3988.67±4.51
Obstructive and restrictive88.00±17.7688.20±15.9287.42±14.86
Spirometry and plethysmographyNormal59.10±10.7365.00±11.4561.63±10.30
Obstructive50.80±42.0435.70±29.29Undefined
Restrictive62.50±11.0359.20±12.2160.45±10.64
Obstructive and restrictive38.20±29.7259.60±44.28Undefined
SpirometryNormal60.70±10.1370.30±10.9964.59±7.92
Obstructive51.50±34.1649.00±36.74Undefined
Restrictive59.80±11.6749.50±17.8953.03±14.47
Obstructive and restrictive44.00±24.7659.00±41.55Undefined
  • Mean±SD for the test sets of 10 experimental runs are shown. In these experiments, data from the initial 477 patients were used for training (80%), validation (10%) and test (10%) sets. For each run, the model randomly selected one test per patient. Biometric- variables are sex, age, height. Spirometry metrics included are FVC, FEV1, FEF25–75 and plethysmography metrics are RV and TLC. Undefined F1 score indicates that either precision, recall or both for that particular class was equal to zero for at least 1 of the 10 experimental runs.

  • FEF, forced mid-expiratory flow; FEV1, forced expiratory volume in 1 s; FVC, forced vital capacity; RV, residual volume; TLC, total lung capacity.