Annals of the Academy of Romanian Scientists Series on Engineering Sciences ISSN 2066-8570 Volume 11, Number 1/2019 51 PREDICTING THE FUNCTIONAL RESULT OF TYMPANOPLASTY WITH THE AID OF NEURONAL NETWORKS USING A REDUCED DATA SET Oana SANDU 1 , Daniel-Petru GHENCEA 2 , Miron ZAPCIU 3 Caius-CodruŃ SARAFOLEANU 4 Rezumat. AplicaŃiile inteligenŃei artificiale îi fac simŃite prezenŃa, în ultimii ani, în foarte multe domenii, inclusiv în medicină. Întrucât medicina este un domeniu cu un nivel de imprevizibilitate mare, metodele tradiŃionale de statistică medicală sunt insuficiente. AplicaŃiile inteligenŃei artificiale, aa cum sunt reŃelele neuronale artificiale, logica fuzzy sau algoritmul genetic, permit crearea de modele inteligente care pot predicŃiona răspunsul pacientului la tratament, sau pot determina riscul la care pacientul este expus pentru o anumită patologie. Lucrarea are scopul de a prezenta un model de reŃele neuronale artificiale pentru predicŃia rezultatului funcŃional după o intervenŃie chirurgicală asupra membranei timpanice (timpanoplastie). Metodologia aplicată în această cercetare are la bază un set redus de date de intrare, una/două intrări în stratul de intrare i ieiri multiple în stratul de ieire. Abstract. Applications of artificial intelligence have made their presence felt over the last few years in many areas, including medicine. Since medicine is a field of high unpredictability, traditional methods of medical statistics are insufficient. Artificial intelligence applications, such as artificial neural networks, fuzzy logic, or genetic algorithm, allow intelligent models to predict the patient's response to treatment or to determine the risk to which the patient is exposed for a particular pathology. Our study aims to present a model of artificial neural networks to predict the functional outcome after surgery on the tympanic membrane (tympanoplasty). The methodology applied in this research is based on a reduced set of input data, one / two entries in the input layer and multiple outputs to the output layer. Keywords: tympanoplasty, artificial neural network, prediction 1. Introduction Computer science changed completely our life in the last years. Computers are used in a diversity of disciplines such as: aerospace, biology, chemistry, physics, 1 PhD Student, Carol Davila University of Medicine and Pharmacy, Bucharest, e-mail: oanastelea@yahoo.com. 2 PhD. Eng.-Ec., University Politehnica of Bucharest, Romania, e-mail: daniel.ghencea@blackseasuppliers.ro 3 Prof. PhD. Eng., University Politehnica of Bucharest, Corresponding member of Academy of Romanian Scientists, miron.zapciu@upb.ro 4 Prof. PhD. Dr., Carol Davila University of Medicine and Pharmacy, Bucharest, e-mail: csarafoleanu@gmail.com