Table 7 Six four-class classifications on AZH dataset.
From: Multi-modal wound classification using wound image and location by deep neural network
Classifications | BG–N–D–V | BG–N–P–V | BG–N–S–V | BG–N–D–P | BG–N–D–S | BG–N–P–S | |
|---|---|---|---|---|---|---|---|
Input | Model | Accuracy (%) | |||||
Location | MLP | 76.58 | 73.29 | 77.27 | 65.38 | 71.74 | 69.04 |
LSTM | 78.48 | 76.03 | 83.12 | 64.62 | 73.91 | 67.46 | |
Image | VGG16 | 93.67 | 89.73 | 87.66 | 82.31 | 77.54 | 83.33 |
VGG19 | 89.87 | 86.99 | 88.31 | 80.00 | 81.88 | 83.33 | |
Image + location | VGG16 + MLP | 94.30 | 91.78 | 94.16 | 86.15 | 86.96 | 85.71 |
VGG19 + MLP | 95.57 | 91.78 | 92.86 | 86.92 | 91.30 | 81.75 | |
VGG16 + LSTM | 89.87 | 92.47 | 90.91 | 86.15 | 84.78 | 83.33 | |
VGG19 + LSTM | 94.30 | 89.04 | 88.89 | 89.23 | 85.51 | 83.33 | |
