关键词: 情绪识别, 社交焦虑, 社会意图理解, 生物运动, 脑影像, 认知计算

Abstract:

Social anxiety disorder (SAD) is among the most common anxiety disorders. SAD is marked by overwhelming fear and avoidance of social scenarios, which debilitates patients’ daily function. Due to the heterogeneous and co-morbid nature of psychiatric disorders, traditional clinical diagnosis methods based on subjective reports and guidelines of DSM and ICD are facing serious challenges, such as misdiagnosis and underdiagnosis. Hence, research is urgently in need to promote the understanding of the psychological and neurobiological mechanisms underlying the clinical symptoms of psychiatric disorders, and to also promote the use of objective biomarkers, such as behavioral and brain activity patterns, to diagnose and predict psychiatric disorders. Nowadays, mood and anxiety disorders are among the most prevalent mental illnesses worldwide that lead to serious outcomes. SAD lies at the intersection of mood and anxiety disorders, offering an opportunity to unfold the mechanisms underlying comorbid mental disorders. SAD is closely associated with abnormal functioning of social cognition. With negative cognitive biases being the representative characteristic, subjects with SAD may demonstrate deficits in both emotional processing and intention inference in social contexts. However, existing evidence in the field cannot readily unify the two important perspectives, emotional processing and social intention inference underlying SAD. The field also lacks effective predictive models of SAD clinical symptoms based on multi-dimensional neurobiological data.
Given these challenges, the current project aims to systematically investigate the cognitive and neural mechanisms underlying emotion processing and intention inference in subjects with high social anxiety traits. Based on the classic biological motion paradigm, we will use a combination of behavioral experiments, functional magnetic resonance imaging (fMRI), clinical measurements of self-report questionnaires, computational modeling, and machine learning algorithms to investigate the negative cognitive bias in SAD. We aim to reveal the underlying unique and shared cognitive neural mechanisms of emotional processing and intention inference, also to establish predictive models of SAD clinical symptoms based on multi-modal data. The current project consists of three experiments. Experiment one will systematically examine the emotional processing of SAD through an emotion judgment task, based on tasks of both the classic emotional biological motion and facial expression recognition. Experiment two will investigate behavioral characteristics and neural mechanisms of SAD in social intention inference through a social interaction judgment task, based on dyad biological motions and facial expression recognition. Biological motions are selected as one of the main testing stimuli in the current study, because they contain rich information of both emotions and social intentions and may be an important source of information for socially anxious people to judge the emotions and intentions of others given their avoidance of facial areas. Therefore, integrating experimental paradigms of biological motions and facial expressions can facilitate the examination of the cognitive biases of emotion recognition and intention inference in social anxiety. Experiment three will integrate the behavioral and neuroimaging data in the previous two experiments to investigate the shared mechanisms of cognitive bias in emotion processing and social intention inference. We aim to examine the link between multimodal data and to investigate the corresponding mechanisms of SAD subtypes and build predictive models.
The project has the prospect to reveal the psychopathology underlying SAD, as well as to examine the association between behavioral and neuroimaging data underlying mental disorders. The current study also has the promise to reveal the role of multimodal data for objective classification and prediction of clinical symptoms. We aim to promote objective classification and prediction of mental disorders based on multimodal data. The efforts may facilitate the realization of the “Health China 2030” plan, which proposed the goal stated as “by 2030, the level of prevention and treatment of common mental disorders and identification of psychological and behavioral problems will be significantly improved”.

Key words: emotion perception, social anxiety, social intention inference, biological motion, brain imaging, cognitive modeling