Table 4 DL (Yolo) risk score model for prognosis and the associated DL features.
From: Radiomics-guided deep neural networks stratify lung adenocarcinoma prognosis from CT scans
Prognostic DL feature | Category | Radiomics feature | Correlation |
|---|---|---|---|
Yolo_Latent_1145 | |||
| Ā | Histogram | Entropy | 0.4559 |
| Ā | Histogram | Uniformity | ā0.4188 |
| Ā | GLCM | Difference average | 0.481 |
| Ā | GLCM | Difference entropy | 0.493 |
| Ā | GLCM | Inverse difference | ā0.5057 |
| Ā | GLCM | IDM | ā0.4893 |
| Ā | GLCM | IDMN | ā0.4025 |
| Ā | GLCM | IDN | ā0.4966 |
| Ā | GLCM | IMC2 | 0.4124 |
| Ā | GLCM | Inverse variance | ā0.493 |
| Ā | GLCM | Sum entropy | 0.4148 |
| Ā | GLSZM | Zone percentage | 0.4718 |
| Ā | Margin | CDF slope skewness | ā0.4467 |
| Ā | Margin | CDF slope kurtosis | ā0.432 |
Yolo_Latent_1161 | |||
| Ā | Shape | Maximum 2D diameter | ā0.4104 |
| Ā | Histogram | Entropy | 0.5011 |
| Ā | Histogram | Uniformity | ā0.4517 |
| Ā | GLCM | Contrast | 0.4523 |
| Ā | GLCM | Difference average | 0.5585 |
| Ā | GLCM | Difference entropy | 0.5581 |
| Ā | GLCM | Inverse difference | ā0.5742 |
| Ā | GLCM | IDM | ā0.5538 |
| Ā | GLCM | IDMN | ā0.4748 |
| Ā | GLCM | IDN | ā0.5743 |
| Ā | GLCM | IMC1 | ā0.4199 |
| Ā | GLCM | IMC2 | 0.4441 |
| Ā | GLCM | Inverse variance | ā0.5601 |
| Ā | GLCM | Sum entropy | 0.4551 |
| Ā | GLSZM | Zone percentage | 0.5417 |
| Ā | Margin | CDF slope skewness | ā0.4905 |
| Ā | Margin | CDF slope kurtosis | ā0.4665 |
Yolo_Latent_22664 | |||
| Ā | GLCM | IDN | 0.4009 |
