Facial Affect Recognition by Patients with Schizophrenia Using Human Avatars

Area of Science:

• Psychiatry
• Cognitive Neuroscience
• Computer Science

Background:

• Schizophrenia impairs facial emotion recognition, affecting social functioning.
• Traditional static stimuli lack real-world dynamism, limiting assessment.
• Dynamic virtual humans offer a novel approach to study facial affect recognition.

Purpose of the Study:

• Evaluate facial emotion recognition in schizophrenia using novel dynamic virtual humans.
• Compare recognition abilities between schizophrenia patients and healthy controls.
• Investigate the influence of stimulus dynamism and viewing angle on recognition.

Main Methods:

• Study included 56 stable schizophrenia patients and 56 healthy controls.
• Utilized a new set of dynamic virtual human facial expressions.
• Assessed recognition rates across different expression dynamisms and viewing angles.

Main Results:

• Schizophrenia patients (71.6% accuracy) significantly underperformed healthy controls (90.0% accuracy) in facial emotion recognition.
• More dynamic expressions and frontal views improved recognition in both groups.
• A trend indicated reduced recognition with higher Positive and Negative Syndrome Scale (PANSS) scores, significant for negative symptoms.

Conclusions:

• Dynamic virtual humans effectively demonstrate facial affect recognition deficits in schizophrenia.
• Stimulus dynamism and viewing angle are critical factors in emotion recognition.
• Findings inform the design of attention-maintaining training interventions for schizophrenia patients.