THE USE OF ARTIFICIAL-NEURAL-NETWORKS METHODOLOGY IN THE ASSESSMENT OF VULNERABILITY TO HEROIN USE AMONG ARMY-CORPS SOLDIERS - A PRELIMINARY-STUDY OF 170 CASES INSIDE THE MILITARY-HOSPITAL-OF-LEGAL-MEDICINE-OF-VERONA

This article describes a preliminary study of screening/diagnostic ins truments for prediction for large-scale application in the military fi eld at the Neuropsychiatric Department of the Military Hospital of Leg al Medicine of Verona and for the prevention of self-destructive behav iors, particularly through the use of drugs. 170 subjects divided into three subsamples were examined. The first subsample was characterized by a strong tendency towards normalcy, the second by a strong tendenc y towards pathology, and the third by a great variety of expressions o f psychological and social problems, which were not necessarily relate d to drug use. These subjects were administered a questionnaire design ed according to Squashing Theory principles (Buscema, 1994a). Answers were processed by an Artificial Neural Network created by Semeion in R ome (Buscema, 1996) and were compared with a standard clinical psychia tric assessment report and with the results of psychodiagnostic tests. Results document ANNs' remarkable ability to recognize subjects with declared, in exordium and ''at risk'' pathological behaviors. Blind re sults on learning and trial samples show a very high predictive capaci ty (over 90%). A comparison with the examined subjects' clinical repor t and the results of the first follow-up also document very high agree ments. The broad variation of answers obtained in the third subsample allows further methodological reflections on the contribution of Artif icial Neural Networks and Squashing Theory to the study of deviance, f or both sociologists and clinicians, and not only for those in the fie ld of drug addiction.