Skip to main content
. 2022 Nov 5;12(11):2702. doi: 10.3390/diagnostics12112702

Table 6.

Performance Evaluation of all models.

Class Precision
%
Sensitivity
%
F1 Score
%
TP FP FN TN Accuracy
%
Loss
Thin Thick Thin Thick Thin Thick Thin Thick Thin Thick Thin Thick Thin Thick Thin Thick Thin Thick
Proposed Model Inf. 95.00 96.00 97.00 98.00 96.00 97.00 4044 2918 197 132 131 50 3896 2900 96.03 96.97 0.12 0.08
Uninf. 97.00 98.00 95.00 96.00 96.00 97.00
VGG16 Inf. 76.00 90.00 85.00 93.00 80.00 91.00 3807 2256 434 794 273 454 3754 2496 91.45 79.20 0.21 0.45
Uninf. 83.00 93.00 74.00 90.00 78.00 92.00
ResNet50 Inf. 62.00 67.00 80.00 75.00 70.00 71.00 2224 1974 2017 1076 799 726 3228 2224 65.94 69.97 0.62 0.57
Uninf. 74.00 73.00 52.00 65.00 61.00 69.00
InceptionV3 Inf. 75.00 76.00 77.00 93.00 76.00 84.00 3064 2297 1177 753 289 671 3738 2279 82.27 76.27 0.39 0.48
Uninf. 77.00 91.00 75.00 72.00 76.00 81.00