Table 3 18F-FDG PET and structural MR-based grey matter volume (GMV) classification accuracy based on the discriminative VOI-based analysis (A) and visual read analysis of two blinded experienced observers (B)
Group | Predicted Group Membership | Total | ||
|---|---|---|---|---|
CON | VSS | |||
(A) Discriminant analysis of VOI-based 18F-FDG PET and MR GMV, after leave-one-out cross-validation | ||||
FDG | ||||
Count | CON | 15 | 0 | 15 |
VSS | 0 | 7 | 7 | |
% | CON | 100 | 0 | 100 |
VSS | 0 | 100 | 100 | |
GMV | ||||
Count | CON | 14 | 1 | 15 |
VSS | 1 | 6 | 7 | |
% | CON | 93 | 7 | 100 |
VSS | 14 | 86 | 100 | |
FDG + GMV | ||||
Count | CON | 15 | 0 | 15 |
VSS | 1 | 6 | 7 | |
% | CON | 100 | 0 | 100 |
VSS | 14 | 86 | 100 | |
Final diagnosis | PET “VSS pattern” | PET normal | ||
|---|---|---|---|---|
Observer 1 | Observer 2 | Observer 1 | Observer 2 | |
(B) Visual classification using 18F-FDG PET | ||||
CON | 0 | 6 | 15 | 9 |
VSS | 5 | 1 | 2 | 6 |