The diffusion model provides a holistic bidirectional context, enabling the system to defer premature judgments during progressive disclosure and graph-based retrieval captures the structured interconnections of clinical knowledge.
Abstract
Mental health disorders affect hundreds of millions of people around the world, yet access to professional counseling remains severely limited. AI-powered dialogue systems offer a scalable alternative, but existing models face two fundamental challenges. First, they lack the bidirectional understanding needed to capture the layered nature of emotional expression, particularly in cases of progressive disclosure, where clients often present symptoms at the surface-level while concealing deeper trauma. Autoregressive (AR) models process information sequentially and cannot revise early interpretations when new evidence emerges later in the conversation. Second, they fail to effectively incorporate the relational knowledge that underlies clinical reasoning. In this paper, we propose \textbf{BiGraph-Diffuse}, the first large-scale diffusion language model tailored for the counseling domain. We further introduce \textbf{BiGraph-RAG}, a relation-free graph-structured retrieval strategy that relies only on lightweight entity extraction and semantic linking. This design preserves inferential pathways from observable symptoms to potential underlying causes, while incurring zero LLM token cost during indexing. Importantly, these two modules are not merely combined but mutually reinforcing. The diffusion model provides a holistic bidirectional context, enabling the system to defer premature judgments during progressive disclosure. Meanwhile, graph-based retrieval captures the structured interconnections of clinical knowledge. Extensive experiments demonstrate the effectiveness of BiGraph-Diffuse, and we further provide a solid theoretical analysis to support its design.
It is demonstrated that integrating clinically grounded retrieval-augmented reasoning with zero-shot prompting significantly enhances the Accuracy and reliability of mental health disorder detection from social media text.
Rajesh Singh Thakur, T. P. Sahu, Gurudatta Verma· International Journal of Mac...· 0 citations
The global burden of mental health challenges disproportionately affects individuals with disabilities. While Artificial Intelligence (AI) presents transformative, accessible interventions, the conceptual structure of this intersection remains fragmented. This study mapped the intellectual landscape of AI-driven mental...
Mohamad Saripudin, A. Badrujaman, Karsih et al.· International Conference on...· 0 citations
Psychotherapy summarisation poses unique challenges due to the emotional nuance, relational complexity, and non-linear discourse of therapeutic dialogues. Existing large language models (LLMs), even when fine-tuned, often fail to capture these structures, leading to clinically incomplete or incoherent summaries. In thi...
P. K. Adhikary, Ankita Mukherjee, Koushik Sinha Deb et al.· IEEE journal of biomedical a...· 0 citations
The findings demonstrate that a leakage-safe evaluation protocol is essential for producing credible results in mental health classification from long social media posts and show that domain-informed density extraction provides a robust and practical text representation, with its greatest benefits emerging when models...
Rafat Hammad, Suha Rababah· Frontiers in Artificial Inte...· 0 citations
MASGR (Multi-Agent Structured Graph Reasoning), a framework that treats referral not as a classification task but as a structured graph construction problem, significantly outperforms state-of-the-art LLMs and existing multi-agent systems, particularly in complex cases requiring the balancing of chronic disease managem...