Probabilistic Textual Time Series Depression Detection
This work addresses the need for accurate and interpretable depression detection for clinical decision support, though it appears incremental as it builds on existing methods like LSTMs and attention with probabilistic extensions.
The paper tackles the problem of predicting depression severity from clinical interviews by proposing PTTSD, a probabilistic textual time series framework that models uncertainty over time, achieving state-of-the-art performance with MAE scores of 3.85 on E-DAIC and 3.55 on DAIC-WOZ datasets.
Accurate and interpretable predictions of depression severity are essential for clinical decision support, yet existing models often lack uncertainty estimates and temporal modeling. We propose PTTSD, a Probabilistic Textual Time Series Depression Detection framework that predicts PHQ-8 scores from utterance-level clinical interviews while modeling uncertainty over time. PTTSD includes sequence-to-sequence and sequence-to-one variants, both combining bidirectional LSTMs, self-attention, and residual connections with Gaussian or Student-t output heads trained via negative log-likelihood. Evaluated on E-DAIC and DAIC-WOZ, PTTSD achieves state-of-the-art performance among text-only systems (e.g., MAE = 3.85 on E-DAIC, 3.55 on DAIC) and produces well-calibrated prediction intervals. Ablations confirm the value of attention and probabilistic modeling, while comparisons with MentalBERT establish generality. A three-part calibration analysis and qualitative case studies further highlight the interpretability and clinical relevance of uncertainty-aware forecasting.