Digital Phenotyping for Spinal Cord Injury: Smartphone-based monitors for clinical utility

Spinal Cord injury (SCI) significantly affects all parts of life, and mental illness and social isolation are common and often undetected after discharge from traditional care. Mobile health and sensor monitoring have emerged as convenient and beneficial supplements to clinical care, even more so with restricted in-person health care during COVID-19. We apply these in SCI to collect and analyze in-situ active self-report as well as passive sensor data from personal smartphones to infer results and correlations between their psychosocial and physical well-being. We have applied Autoregressive Integrated Moving Average (ARIMA) to understand time dependent relationships between depression severity, social interaction, and community mobility, and explored clustering analysis and parallel predictive models to inform just-in-time adaptive interventions. Preliminary analyses suggest that smartphones, as a symptom monitoring tool and to deliver an in-situ individualized intervention have potential to positively impact depression severity and community participation after SCI.