16892
Self-Adjusting Biofeedback with a Dynamic Feedback Signal Set (DyFSS)

Friday, May 16, 2014
Meeting Room A601 & A602 (Marriott Marquis Atlanta)
L. I. Sugarman1, B. L. Garrison2, A. E. Hope3, S. Jacobs4, A. J. Glade5 and K. L. Williford5, (1)Rochester Institute of Technology, Pittsford, NY, (2)153 Lomb Memorial Dr, Rochester Institute of Technology, Rochester, NY, (3)Center for Applied Psychophysiology and Self-regulation, Rochester Institute of Technology, Rochester, NY, (4)Interactive Games and Media, Rochester Institute of Technology, Rochester, NY, (5)Rochester Institute of Technology, Rochester, NY
Background:  The diversity found in Autism Spectrum Disorder (ASD) may also be reflected in the autonomic profile of affected individuals, yet there is a paucity of normative data for how their autonomic functioning may differ. It is known that anxiety is highly prevalent in children with ASD, (Baron, Groden, Groden, & Lipsitt, 2006; Kinsbourne, 2011). Autonomic Biofeedback Training (ABT) is a promising treatment for managing anxiety and ASD symptoms more generally (Sugarman, Garrison, Williford, 2013; Yucha & Montgomery, 2008). However, a clinician looking to use ABT for youth with ASD cannot readily predict which autonomic proxies are the most discernable and controllable for a specific user. Clinical encounters would benefit from software that tunes a combination of sensor signals to the best abilities and needs of each individual patient. We are developing a Dynamic Feedback Signal Set (DyFSS), a strength-based, self-customizing algorithm. It uses four measures of autonomic function: skin conductance, skin temperature, respiration rate, and low-frequency heart rate variability. By creating individualized and intuitive software, ABT can be refined to address the autonomic heterogeneity of youth with ASD, ease the work of the clinician, and create the potential for integration of ABF into interactive games and media.

Objectives: To test the feasibility and receive input from potential users with ASD regarding (1) a novel biofeedback algorithm that combines four autonomic signals for a unified display interface; (2) a 5-session weekly ABT protocol employing this algorithm; and, (3) parent behavioral observation as an outcome measure for this course of ABT. Presentation includes a technological demonstration of the current DyFSS prototype.

Methods: The first version of the DyFSS was tested over 5 weekly sessions of ABT with 10 youth diagnosed by community physicians with ASD. User preferences were obtained by asking participants for direct input and assessed qualitatively. Daily behavior tracking by parents tested for change in ASD symptoms. A questionnaire completed by users at the close of the final session was assessed qualitatively to assess their overall experience.

Results: Initial reactions show that many children are interested in learning more about biofeedback as well as the technology and physiology underlying the process. Some report their use of the skills learned during ABT to cope with stressors in school and at home between sessions. Analysis of parental observation is pending at this time. Areas to improve the current version of the DyFSS include the setup of the physical sensors, customizability of the graphical user interface (GUI), and follow-up to determine whether children continue to use their skills after sessions end.

Conclusions: Youth with ASD are readily engaged through technological interventions such as autonomic biofeedback. It is an effective way to draw interest toward therapy and increase understanding of their physiological processes. It may also decrease their anxiety and associated symptoms. Further refinements of the DyFSS based on input from youth with ASD will improve the relevance of the software in clinical practice and its potential integration into interactive games and media.