Mental-disorder detection using chaos and nonlinear dynamical analysis of photoplethysmographic signals

• Chaos and nonlinear dynamical analysis are applied for mental-disorder detection.
• Experimental results show significant detection improvement with feature synergy.
• Proposed approach is effective for analysis of photoplethysmographic signals.
• Proposed approach is promising for developing automated mental-health systems.

Mental disorder can be defined as a psychological disturbance of thought or emotion. In particular, depression is a mental disease which can ultimately lead to death from suicide. If depression is identified, it can be treated with medication and psychotherapy. However, the diagnosis of depression is difficult and there are currently no any quick and reliable medical tests to detect if someone is depressed. This is because the exact cause of depression is still unknown given the belief that depression results in chemical brain changes, genetic disorder, stress, or the combination of these problems. Photoplethysmography has recently been realized as a non-invasive optical technique that can give new insights into the physiology and pathophysiology of the central and peripheral nervous systems. We present in this paper an automated mental-disorder detection approach in a general sense based on a novel synergy of chaos and nonlinear dynamical methods for the analysis of photoplethysmographic finger pulse waves of mental and control subjects. Such an approach can be applied for automated detection of depression as a special case. Because of the computational effectiveness of the studied methods and low cost of generation of the physiological signals, the proposed automated detection of mental illness is feasible for real-life applications including self-assessment, self-monitoring, and computerized health care.