Validation of Machine Learning-Based Assessment of Major Depressive Disorder from Paralinguistic Speech Characteristics in Routine Care

Table 4

Accuracy, sensitivity, and specificity of classifications based on the PSP system and based on cut-off scores from depression scales.

Measure	Balanced accuracy	Sensitivity	Specificity
PSP-based	0.66	0.70	0.62
HRSD score	0.73	0.83	0.62
QIDS-C score	0.74	0.87	0.61
PHQ-9 score	0.73	0.66	0.80

Interview and questionnaire data across time points. PSP = paralinguistic speech processing; HRSD = Hamilton Rating Scale for Depression; QIDS-C = Quick Inventory of Depressive Symptomatology–Clinician-rated; PHQ-9 = depression module of the Patient Health Questionnaire.