Multimodal Conversational Technology for Remote Assessment of Symptom Severity in People with Schizophrenia

Abstract

The causes of schizophrenia are complex and largely unknown. Current treatments focus on the management of positive or psychotic symptoms. However, negative symptoms like lack of facial and verbal expression, monotone speech, social withdrawal and isolation affect people with schizophrenia on a day-to-day basis and are harder to recognize and interpret. A shortage of mental health services in parts of the world makes it difficult to provide care and support to people with schizophrenia. Additionally, the COVID-19 pandemic exposed the pressing need for clinicians to be able to monitor treatment and symptoms in psychiatric disorders remotely. We present an exploratory study highlighting the feasibility of a novel multimodal conversational platform with real-time automated extraction of speech and facial metrics to assess symptom severity in people with schizophrenia.

Publication
In Proceedings of the Society for Neuroscience (SfN) Annual Meeting