发表机构
The Chinese University of Hong Kong, Shenzhen; Shenzhen Loop Area Institute（香港中文大学（深圳）; 深圳河套学院）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
本研究提出顺序LoRA跨尺度迁移方法，从英文DAIC-WOZ（PHQ-8）迁移到中文PDCH（HAMD-17），在数据稀缺下提升抑郁评分预测，Qwen3-1.7B取得最佳MAE 4.38。
AI 中文摘要
本研究针对临床访谈转录文本在数据稀缺条件下的连续抑郁严重程度评分预测问题。我们提出了一种用于跨尺度迁移的顺序低秩适应（LoRA）协议：一个带有有界回归头的Qwen3骨干模型首先在英文DAIC-WOZ数据集（189次虚拟人中介会话，PHQ-8）上进行微调，随后该适配器初始化在中文PDCH数据集（100次真实临床咨询，HAMD-17）上的微调，其中重新初始化的、尺度特定的回归头预测临床医生分配的评分。所有配置均使用患者级分层5折、2次重复交叉验证。在数据稀缺的HAMD-17目标上，顺序协议在0.6B和1.7B骨干模型上均取得了最佳的点估计MAE、RMSE和宏F1，优于仅目标训练和非LLM基线——Qwen3-0.6B达到4.96/6.59/0.36，Qwen3-1.7B达到4.38/5.62/0.46。消融实验表明，正确对齐的源监督给出最佳点估计（无监督暴露和打乱标签对照也显示部分增益），原生中文目标输入优于机器翻译的英文输入，且反向顺序在运行间方差内没有明显增益。本研究是探索性的、单中心内部评估：它不建立筛查或诊断效用，也不单独识别尺度、语言或范式转变的贡献。据我们所知，尚无先前研究评估这种特定的DAIC-WOZ到PDCH顺序迁移设置。
英文摘要
This work addresses continuous depression-severity score prediction from clinical interview transcripts under data scarcity. We propose a sequential low-rank adaptation (LoRA) protocol for cross-scale transfer: a Qwen3 backbone with a bounded regression head is first fine-tuned on the English DAIC-WOZ dataset (189 avatar-mediated sessions, PHQ-8), and the adapter then initializes fine-tuning on the Chinese PDCH dataset (100 real clinical consultations, HAMD-17), where a reinitialised, scale-specific head predicts the clinician-assigned score. All configurations use patient-level stratified 5-fold, 2-repeat cross-validation. On the data-scarce HAMD-17 target, the sequential protocol attains the best point-estimate MAE , RMSE, and macro-$F_1$ on both 0.6B and 1.7B backbones, outperforming target-only training and non-LLM baselines---4.96/6.59/0.36 with Qwen3-0.6B and 4.38/5.62/0.46 with Qwen3-1.7B. Ablations suggest that correctly aligned source supervision gives the best point estimates (unsupervised exposure and shuffled-label controls also show partial gains), that native-Chinese target input outperforms machine-translated English input, and that the reversed order yields no clear gain within run-to-run variance. The study is an exploratory, single-site internal evaluation: it does not establish screening or diagnostic utility, nor separately identify the contribution of the scale, language, or paradigm shifts. To our knowledge, no prior study evaluates this specific DAIC-WOZ-to-PDCH sequential transfer setting.
Commentspreprint to ICASSP 2027
