MEDIÇÃO DA INTENSIDADE DA DOR COM A UTILIZAÇÃO DE REDES BAYESIANAS

Abstract

Pain is subjective and personal, and knowing the intensity of pain that a person is feeling enables appropriate medical interventions to be carried out, seeking to avoid excessive medication on one hand, or leaving the patient in pain on the other. This dissertation presents a proposal for measuring the intensity of pain in humans, through the development of a computer system based on Bayesian networks that are capable of analyzing data from biomedical signals and other variables related to pain. With the certainty of the occurrence of any one variable, the computer system performs probabilistic inference to measure the intensity of the pain that a person is feeling. In this study, pain was simulated by immersing the subjects arms in ice, as they communicated verbally, and analyzing the biomedical changes (blood pressure, and oxygenation, temperature, heart rate and respiratory rate) and emotional changes (anxiety, pallor, muscle tension, body position and facial expression) that occurred as a result of the pain. Based on the comparison between the responses in the analogue pain scales and the results presented, causal network modeling and probabilistic values were established. The results obtained demonstrate the efficiency of the Bayesian network to infer probabilistically the intensity of pain, but in order to further refine the results, a sample with a larger number of participants is needed