Multimodal Physiological Markers of Affect for Automated Depression Analysis

Depression is the most common mental health disorder with high societal costs. It is the most common cause of disability worldwide and a major risk factor for suicide. Individuals with depression experience persistent and pervasive low mood and anhedonia. Despite high prevalence and social impact, there are no laboratory-based assessment tools that can objectively measure severity of depression and its recovery. Typically, depression is assessed by clinicians using questionnaires and interviews, risking a range of subjective biases and misinterpretations.

Our main ambition is to develop assistive technologies supporting clinicians in diagnosis and monitoring the progress of depression in objective and automated ways. To this line, this project aims to identify robust and reliable multimodal physiological markers of affective states and investigate their effectiveness in detecting and monitoring depression in real-time. Thus, we will develop advanced multimodal signal processing pipelines and machine learning techniques for mapping physiological signals (i.e. brain, heart and skin electrodermal signals) recorded during passive and active mood challenging tasks onto reliable diagnostic and prognostic metrics of depression. In addition, the identified multimodal physiological markers can be potentially used for improving the treatment of the patients with depression using biofeedback methods.

The prestigious A*STAR-Sheffield scheme gives the PhD student the unique opportunity to carry out part of their studies in Singapore as part of a partnership agreement with the Agency for Science, Technology and Research (A*STAR). The prospective PhD student will receive half of their training (2 years) in Singapore. Supervision on this project will be provided by Dr Mahnaz Arvaneh from Department of Automatic Control and Systems Engineering and co-supervision of Dr Liat Levita from School of Psychology. The project will run in partnership with the Institute for Infocomm Research (I2R) in Singapore with co-supervision of Dr Haihong Zhang.