Clinical Outcome Prediction in Aneurysmal Subarachnoid Hemorrhage Using Bayesian Neural Networks with Fuzzy Logic Inferences

Table 4

Results of Artificial Neural Networks reveal normalized importance values of predictor variables in aneurysmal subarachnoid hemorrhage.

Artificial neural networks; independent variable	Type of prognostic factor	Importance
Age	Demographic	0.111
Second stroke	Neurologic	0.081
Myocardial infarction	Systemic	0.075
Temperature	Systemic	0.061
Mean arterial pressure	Systemic	0.054
Neurological grade	Neurologic	0.048
Ruptured aneurysm size	Neurologic	0.039
Diabetes mellitus	Systemic	0.037
Angina	Systemic	0.034
SAH clot thickness	Neurologic	0.033
Lung edema	Systemic	0.032
Admission angiographic vasospasm	Neurologic	0.029
Previous subarachnoid hemorrhage	Neurologic	0.028
Vasospasm day	Neurologic	0.028
Cerebral edema	Neurologic	0.028
Vasospasm during treatment	Neurologic	0.027
Aneurysm location	Neurologic	0.025
Time to treatment	Demographic	0.025
Normal motor response	Neurologic	0.024
Intracerebral hematoma	Neurologic	0.022
Normal speech	Neurologic	0.021
Day-8 temperature	Systemic	0.021
Gender	Demographic	0.020
Eye opening	Neurologic	0.018
Migraine history	Neurologic	0.015
Intraventricular hemorrhage	Neurologic	0.015
Hypertensive history	Systemic	0.014
Anticoagulant use	Systemic	0.014
Seizures	Neurologic	0.013
Hydrocephalus	Neurologic	0.012