Investigation of the Accuracy of Depression Prediction Based on Speech Processing

The present study investigates the accuracy of prediction of depression based on speech processing. Depression is one of the most widespread psychiatric disorders, but early detection of depression is difficult. The Beck Depression Inventory II (BDI) is a self-assessment questionnaire and can accurately predict the severity of depression. BDI is most often used for early detection. There is no known objective biomarker for depression, but the state alters the speech of the individual suffering from it, providing an opportunity for speech-based detection. In the current study, we investigated the accuracy of prediction of depression severity based on speech signal processing, and how accurately it can predict the severity of depression compared to the Beck Depression Inventory.