ORAKEL
Better prediction of relapses in depressive disorders by detecting early warning signs using AI.
Motivation and Goals
Recent advances in AI and machine learning offer promising opportunities to improve early detection of worsening depressive symptoms. Preliminary studies suggest that AI can analyze subtle cues from speech patterns, facial expressions and gestures to detect depressed mood and suicidal crises. For example, depressed people may exhibit changes in the prosody of speech, reduced facial expressions and spontaneous gestures. There is also evidence that vital signs such as heart rate variability and sleep patterns are indicative of a person’s mental state. In our project, we will directly compare how well the assessment of the patient’s state of illness or their risk of relapse succeeds: (a) through the medical consultation (as has been common up to now), (b) through standardized ratings or interviews (as is currently common in psychiatric research), (c) by predicting relapses in depressive disorders through the detection of early warning signs by means of AI (new approach of our project), (d) by combining the aforementioned approaches.
This will allow us to see not only whether AI is in principle capable of detecting early warning signs of depression in a clinical context, but also whether this works better than conventional methods. Camera-based monitoring and AI-driven analyses could then provide real-time feedback for healthcare providers and thus enable earlier interventions. The detection of early warning signs of a relapse using artificial intelligence therefore offers considerable potential for improving the care of patients with depressive disorders. The further development of such technologies can also be a helpful addition, particularly due to the limited time resources available for outpatient care as a result of the shortage of doctors. The addition of AI to analyze speech, facial expressions, gestures and vital signs in the assessment of the course of the disease could help to better manage the outpatient treatment of depressive disorders and sustainably improve the quality of life of those affected.
Funding
This project is supported by the European Regional Development Fund (ERDF) as part of project ORAKEL (grant no. ZS/2023/12/182322), funded by the European Union and the state of Saxony-Anhalt.
