Table 3.
Match of predicted values of individual, multistructural and multimodal study metrics with actual outcomes (Engel class 1 vs. Engel class ≥ 2).
| Criteria | (i) Vol Ratio | (ii) FLAIR | (iii) MSA | (iv) PET | (v) MMM | (vii) DTI Uncertainty | (vii) DTI Connectivity | (viii) fMRI Connectivity |
|---|---|---|---|---|---|---|---|---|
| Engel 1 | 13/17 = 0.76 | 9/13 = 0.69 | 15/17 = 0.88 | 10/12 = 0.83 | 14/17 = 0.82 | 7/10 = 0.70 | 5/9 = 0.56 | 2/2 = 1 |
| Engel ≥2 | 4/7 = 0.57 | 3/7 = 0.43 | 3/7 = 0.43 | 2/3 = 0.67 | 4/7 = 0.57 | 0/3 = 0.00 | 1/3 = 0.33 | - |
| Engel class 1 Outcomes | Engel class ≥ 2 Outcomes | |||||||
| R \ M | L&UL | R&UR | U | L&UL | R&UR | U | ||
| Hippocampal Volume | L | 6 | 1 | 0 | 2 | 2 | 0 | |
| R | 2 | 7 | 1 | 1 | 2 | 0 | ||
| FLAIR | L | 4 | 1 | 1 | 2 | 1 | 1 | |
| R | 1 | 5 | 1 | 1 | 1 | 1 | ||
| MSA | L | 5 | 2 | 0 | 1 | 2 | 0 | |
| R | 0 | 8 | 1 | 1 | 2 | 0 | ||
| PET | L | 4 | 0 | 0 | 1 | 0 | 0 | |
| R | 1 | 6 | 1 | 0 | 1 | 1 | ||
| MMM | L | 5 | 1 | 1 | 2 | 1 | 1 | |
| R | 0 | 9 | 1 | 0 | 2 | 1 | ||
| DTI (u) | L | 5 | 0 | 0 | 0 | 0 | 1 | |
| R | 0 | 1 | 4 | 1 | 0 | 1 | ||
| DTI (c) | L | 2 | 0 | 1 | 0 | 0 | 1 | |
| R | 1 | 3 | 2 | 0 | 0 | 2 | ||
Note: each row in the upper display shows the number of cases predicted in each category leading to the identified outcome so that for Engel class ≥ 2, the number of wrong-side predictions are provided. Comparative tables (Engel class 1 vs. Engel class ≥ 2) show outcomes achieved by resection (R) following standard clinical protocol decision-making (left vs. right) and by model prediction (M) for each attribute, indicating absolute or partial determinacy for the epileptogenic side (L & UL vs. R & UR).