Assessment of emotional states through physiological signals and Its application in music therapy for disabled people

Disabled people have more difficulties for expression and interaction on a non-verbal level because of the physical, sensory or communication difficulties that they usually present. Regarding music therapy, the existing adapted interfaces target disabled people with controlled motor abilities, which still suggests a barrier for the most affected users presenting truly little or uncontrolled movements. In this work, we have developed a device to build an adapted musical instrument through physiological signals. The instrument uses the electrocardiogram (ECG), the electrical activity of the skin (EDA), the respiration signal and the movement of the user to generate music via sonification of the characteristic features of each physiological signal. Furthermore, the ECG and EDA signals have been used to assess the emotional state of the person to provide to the music therapist objective information in real-time about the adaptation of the user to the techniques and interfaces used in their sessions. The adapted instrument has been tested by people with cerebral palsy showing its high degree of adaptability to the user. In four months, an increase in participation of the most affected users in their music therapy sessions has been achieved. In concrete, the results show that the considered users have achieved six of the so-called aids to analysis which are commonly used to evaluate such practices. The results of the assessment of emotional states indicate that the state of a person can be extracted from the ECG in periods of ten seconds while the evolution of the EDA reveals if the person is relaxed or excited. The device has generated a lot of interest among educators since it outperforms the state-of-the-art techniques allowing the integration of the most affected users by eliminating the aforementioned barrier because of non-controlled movements while assessing the emotional state of the person when facing the activity.Postprint (published version