RT Journal Article T1 Fear Detection in Multimodal affective computing: Physiological Signals versus Catecholamine Concentration A1 Gutiérrez Martín, Laura A1 Romero Perales, Elena A1 Sainz de Baranda Andújar, Clara A1 Canabal Benito, Manuel Felipe A1 Rodríguez-Ramos, Gema Esther A1 Toro-Flores, Rafael A1 López-Ongil, Susana A1 López Ongil, Celia AB Affective computing through physiological signals monitoring is currently a hot topic inthe scientific literature, but also in the industry. Many wearable devices are being developed forhealth or wellness tracking during daily life or sports activity. Likewise, other applications are beingproposed for the early detection of risk situations involving sexual or violent aggressions, with theidentification of panic or fear emotions. The use of other sources of information, such as video or audiosignals will make multimodal affective computing a more powerful tool for emotion classification,improving the detection capability. There are other biological elements that have not been exploredyet and that could provide additional information to better disentangle negative emotions, suchas fear or panic. Catecholamines are hormones produced by the adrenal glands, two small glandslocated above the kidneys. These hormones are released in the body in response to physical oremotional stress. The main catecholamines, namely adrenaline, noradrenaline and dopamine havebeen analysed, as well as four physiological variables: skin temperature, electrodermal activity, bloodvolume pulse (to calculate heart rate activity. i.e., beats per minute) and respiration rate. This workpresents a comparison of the results provided by the analysis of physiological signals in reference tocatecholamine, from an experimental task with 21 female volunteers receiving audiovisual stimulithrough an immersive environment in virtual reality. Artificial intelligence algorithms for fearclassification with physiological variables and plasma catecholamine concentration levels have beenproposed and tested. The best results have been obtained with the features extracted from thephysiological variables. Adding catecholamine’s maximum variation during the five minutes afterthe video clip visualization, as well as adding the five measurements (1-min interval) of these levels,are not providing better performance in the classifiers. PB MDPI AG SN 1424-3210 YR 2022 FD 2022-05-26 LK https://hdl.handle.net/10016/34976 UL https://hdl.handle.net/10016/34976 LA eng NO This research has been supported by the Madrid Governement (Comunidad de Madrid,Spain) under the ARTEMISA-UC3M-CM research project (reference 2020/00048/001), the EMPATIACMresearch project (reference Y2018/TCS-5046) and the Multiannual Agreement with UC3M in theline of Excellence of University Professors (EPUC3M26), and in the context of the V PRICIT (RegionalProgramme of Research and Technological Innovation). DS e-Archivo RD 1 sept. 2024