Process Analysis of Depression Expression and Action Changes and Their Psychological Integration Based on Deep Learning Algorithms

This paper explores the relationship between the changes in facial expressions and actions of patients with depression and the outcome of their symptoms and internal psychological integration based on deep learning algorithms. We collected full audio and video materials of 14 cognitive-behavioral therapy sessions from one patient with depression. We used the Hamilton Depression Scale and the Hamilton Anxiety Scale to assess the patient’s condition. We applied deep learning algorithms to analyze the patient’s facial entropy, facial components, and action amplitude. An integrative analysis method evaluated the patient’s problem experience integration sequence. Results: The reduction rates of HAMD and HAMA of the patient before and after treatment were 90.91% and 91.43%, respectively. As the treatment progressed, the patient’s facial entropy showed a fluctuating rise, the probability of depressive expressions overall showed a decreasing trend, the overall amplitude of movements gradually increased, and the volume of speech gradually increased. Applying deep learning algorithms to recognize the expressions and posture evaluations of patients with depression is feasible.