SENSITIVITY PROFILE OF 41 PSYCHIATRIC PARAMETERS DETERMINED BY NEURAL-NETWORK IN RELATION TO 8-WEEK OUTCOME

Neural network processed 41 variables of 289 psychiatric inpatients an d successfully converged (97% correlation between output and target si gnals) on a good 8-week outcome. The sensitivity of each parameter was calculated by ditherring the input signals one by one at 1% signal in crease and computing the corresponding output Fifteen parameters were predictors of good outcome and 17 parameters were predictors of bad ou tcome. The most sensitive good outcomes were ECT, group, behavioral an d dynamic psychotherapeutic approaches, clozapine administration, nora drenergic antidepressant treatment schizopositive symptoms, and being of high socioeconomic level. The most sensitive bad outcomes were long duration of the East in-hospital stay or last remission suffering fro m OCD or resistant depression, being treated with low potency neurolep tic drugs, and having community support. Sensitivity detection by neur al network is a new approach in psychiatry. A 70% concordance level wa s found with outcome measures reported in literature and that have bee n traditionally calculated. The fact that the system converged in spit e of the wide spectrum of input signals means that underlying intercon nections were detected by the neural network.