EmoTrack: Robust Depression Tracking from Counseling Transcripts across Session Regimes

2026-05-21

AI Summary

Existing approaches struggle to robustly track depression severity (PHQ-8) from counseling transcripts under data scarcity or when modeling longitudinal context. To address this, this work proposes EmoTrack, a framework that integrates clinical signals elicited via prompt-engineered large language models with frozen sentence-level semantic embeddings to construct symptom-specific regressors. In multi-session settings, EmoTrack further incorporates a compact cross-session memory mechanism to model symptom evolution. This is the first method evaluated for longitudinal depression tracking under realistic conditions of partial symptom disclosure and cross-session continuity, balancing data efficiency with contextual modeling capacity. Experiments demonstrate that EmoTrack significantly outperforms baselines on both the LongCounsel and DAIC-WOZ datasets, reducing single-session mean absolute error (MAE) by 13.5% and achieving state-of-the-art performance in longitudinal tracking.

Abstract

Text-based counseling is an important interface for AI mental-health support, where transcripts may be used to monitor depression severity and flag sessions requiring timely human review. However, robust PHQ-8 prediction across session regimes remains challenging: fine-tuning-based methods can exploit richer supervision but may generalize poorly under data scarcity, while prompt-based LLM methods are data-efficient but usually treat each transcript holistically and provide limited support for longitudinal context. We study robust depression tracking from counseling transcripts across single-session and multi-session regimes. We introduce LongCounsel, a multi-session counseling dataset with session-level PHQ-8 supervision for evaluating repeated-session tracking under partial symptom disclosure and cross-session continuity. We further propose EmoTrack, a PHQ-8 prediction framework that combines LLM-extracted clinical signals with frozen turn-level semantic embeddings and trains symptom-specific predictors over the resulting transcript representation. When prior sessions are available, EmoTrack can further incorporate them through compact cross-session memory. Experiments on LongCounsel and DAIC-WOZ show that EmoTrack achieves a clear gain on the real single-session benchmark, including a 13.5% relative MAE reduction over the strongest DAIC-WOZ baseline, and remains competitive with the strongest longitudinal baseline on LongCounsel.

Problem

• depression tracking
• counseling transcripts
• PHQ-8 prediction
• session regimes
• longitudinal context

Innovation

• EmoTrack
• depression tracking
• PHQ-8 prediction
• longitudinal context
• LLM-extracted clinical signals