Table 5.
|
Components |
||||||
---|---|---|---|---|---|---|---|
Predictors | 1 | 2 | 3 | 4 | 5 | 6 | 7 |
Age | −0.05 | 0.07 | 0.98 | 0.09 | 0.02 | −0.13 | 0.09 |
GCS | 0.95 | −0.11 | −0.06 | −0.17 | −0.06 | 0.09 | −0.21 |
SAH/IPH volume | −0.05 | 0.02 | 0.02 | 0.05 | 0.99 | 0.04 | 0.06 |
SDH volume | −0.18 | 0.27 | 0.10 | 0.92 | 0.07 | −0.04 | 0.18 |
EDH volume | 0.08 | −0.02 | −0.13 | −0.03 | 0.05 | 0.99 | −0.03 |
Shift (distance) | −0.12 | 0.91 | 0.08 | 0.28 | 0.02 | −0.02 | 0.28 |
Cistern effacement | −0.29 | 0.34 | 0.13 | 0.21 | 0.08 | −0.04 | 0.86 |
Principal component analysis of quantitative computed tomography (qCT) predictors, age and GCS. Thick-lined boxes highlight high loading coefficients ≥0.9. Thin-lined boxes demonstrate intermediate loading coefficients from ∼0.3 up to 0.9. A principal component that is driven purely by a single predictor demonstrates a coefficient of 1.0 for a single predictor, and a coefficient of zero for all other predictors under consideration. Here, components 3, 5, and 6 were each driven almost solely (loading coefficient ≥0.98) by a single feature (age, SAH/IPH volume, and epidural hematoma volume, respectively), indicating that these predictors had little collinearity with each other or with other predictors. In contrast, components 1, 2, 4, and 7 consisted of significant contributions from more than one predictor. Components 2, 4, and 7 demonstrated covariances among features related to intracranial mass effect, including subdural hematoma volume, severity of midline shift, and severity of basal cistern effacement, and Component 1 demonstrated a negative correlation of admission GCS with severity of cistern effacement; both of these result from an entirely data-driven principal component analysis approach, yet are also intuitively satisfying.
qCT, quantitative CT; GCS, Glasgow Coma Scale; SAH, subarachnoid hemorrhage; IPH, intraparenchymal hemorrhage; SDH, subdural hematoma; EDH, epidural hematoma.