ConsultMind:Towards Automated Diagnostic Consultation via Uncertainty-Aware Reasoning

Xiao Sun; Yuming Yang; Yun Chen; Jiang Zhong; Junnan Zhu; Xinyi Jiang; Haoyang Zeng; Ruirui Chen

cs.AISep 25, 2026

ConsultMind:Towards Automated Diagnostic Consultation via Uncertainty-Aware Reasoning

Authors: Xiao Sun, Yuming Yang, Yun Chen, Jiang Zhong, Junnan Zhu, Xinyi Jiang, Haoyang Zeng, Ruirui Chen,

Organizations: College of Computer Science, Chongqing University · College of Computer Science, Hunan University · Institute of Automation, Chinese Academy of Sciences · School of Computer Science and Engineering, University of New South Wales ·

Abstract

Diagnostic consultation is an online sequential decision-making process in which clinicians gather evidence through patient interaction until a diagnosis is sufficiently supported. Automating this process requires adaptive inquiry and interpretable decisions. Bayesian networks offer a natural foundation by updating diagnostic posteriors as evidence accumulates, but their use in open-ended consultation raises two challenges: linking diagnostic hypotheses to potential inquiries and translating evolving posteriors into consultation decisions. We introduce AutoDisym, an automated pipeline that integrates diagnostic knowledge with heterogeneous diagnosis-labeled clinical narratives to construct a Disorder--Symptom Bayesian Network (DSBN). Building on the DSBN, we propose ConsultMind, an uncertainty-aware framework that updates disorder posteriors after each response and uses posterior uncertainty to guide inquiry and diagnosis. We evaluate both methods across psychiatry, respiratory medicine, fever clinics, and three public datasets. The results show that AutoDisym can automatically construct high-quality DSBNs and that ConsultMind consistently improves diagnostic performance and explanation soundness. For example, AutoDisym achieves macro-averaged F1 scores of 81.37 for canonical symptoms and 72.19 for manifestations using GPT-5.6-Sol. ConsultMind improves Top-1 and Top-3 diagnostic accuracy by up to 22.15 and 37.89 percentage points, respectively. Physician evaluation further shows that ConsultMind improves the quality of ranking explanations, differential diagnoses, and diagnosis rationales across LLMs of different scales. This work offers a promising approach to automatic diagnostic consultation.

Figures & tables

Figure 1: Comparison of automated diagnostic consultation methods. Protocol- and LLM-driven systems suffer from rigid inquiry and opaque decisions, respectively, whereas ConsultMind enables adaptive and interpretable consultation through uncertainty-aware reasoning.

Figure 2: Overview of AutoDisym . An automated pipeline that constructs Disorder–Symptom Bayesian Networks (DSBN) from diagnostic knowledge and clinical narratives.

Figure 3: Automated Diagnostic Consultation via ConsultMind . It grounds patient responses into canonical symptom states, updates diagnostic posteriors through the DSBN, and uses posterior uncertainty to guide inquiries until reaching a diagnosis.

	Psychiatry			Respiratory Medicine			Fever Clinic
Configuration	Top-1 $\uparrow$	Top-3 $\uparrow$	MRR $\uparrow$	Top-1 $\uparrow$	Top-3 $\uparrow$	MRR $\uparrow$	Top-1 $\uparrow$	Top-3 $\uparrow$	MRR $\uparrow$
Qwen3-8B
Direct	21.05%	35.79%	0.279	17.42%	31.64%	0.255	14.26%	26.91%	0.222
w/ RAG	28.31% +7.26	51.11% +15.32	0.383 +0.104	22.27% +4.85	42.56% +10.92	0.328 +0.073	19.01% +4.75	36.38% +9.47	0.284 +0.062
w/ ConsultMind	41.80% +20.75	69.84% +34.05	0.538 +0.259	31.28% +13.86	55.91% +24.27	0.438 +0.183	27.84% +13.58	47.95% +21.04	0.376 +0.154
Llama-3.1-8B

Table 1: Comparison of LLMs under direct diagnosis, diagnostic-criteria RAG, and ConsultMind across clinical specialties. Top-

k

denotes diagnostic accuracy; MRR is the mean reciprocal rank of the correct diagnosis.

Base Model	Cov. $\uparrow$	Prec. $\uparrow$	Halluc. $\downarrow$
QwQ-32B	72.40	86.10	7.80
w/ Auditor	88.91 $(+16.51)$	93.27 $(+7.17)$	3.08 $(-4.72)$
DeepSeek-V4-Flash	76.85	88.74	6.42
w/ Auditor	90.32 $(+13.47)$	94.15 $(+5.41)$	2.61 $(-3.81)$
GPT-5.5	82.36	91.52	4.63
w/ Auditor	92.18 $(+9.82)$	95.06 $(+3.54)$	1.94 $(-2.69)$

Table 2: Projected VSP reliability with and without auditing.

Figure 4: Comparison of ConsultMind with medical LLMs.

Model	Top-1 $\uparrow$	Top-3 $\uparrow$	MRR $\uparrow$
MentalHospital Psychiatry
Llama-3.1-8B	16.77%	34.19%	0.249
w/ ConsultMind	33.55% +16.78	51.61% +17.42	0.430 +0.181
Qwen3-32B	27.74%	35.81%	0.325
w/ ConsultMind	40.65% +12.91	60.32% +24.51	0.512 +0.187
GPT-5.6-Sol	29.03%	34.19%	0.329

Table 3: Diagnostic performance of LLMs with and without ConsultMind across datasets.

Figure 5: Physician ratings of diagnostic explanation quality.

Figure 6: Consult comparison with/without ConsultMind .

Qwen3-32B
	Symptoms			Manifestations
Configuration	Precision $\uparrow$	Recall $\uparrow$	F1 $\uparrow$	Precision $\uparrow$	Recall $\uparrow$	F1 $\uparrow$
Direct	52.17	42.46	46.82	40.86	29.36	34.17
w/ RAG	67.69	61.34	64.36	57.89	49.54	53.39
w/ AutoDisym	76.78	72.75	74.71	69.46	63.28	66.23
GPT-5.6-Sol

Table 4: Schema agreement across configurations.

Medical Diagnosis

Medical diagnosis uses clinical data, tests, and images to identify diseases, assess risk, and guide care.

See all 262 papers

Jun 2, 2026

MeDxAgent: Multi-Agent Consultation for Interactive Medical Diagnosis

Large language models (LLMs) are increasingly used for health-related decision support. Yet most evaluations treat diagnosis as a single-shot task with complete information provided upfront, often as a multiple-choice selection. This diverges from clinical practice, where diagnosis is interactive and open-ended, involving sequential hypothesis refinement through targeted questioning. We address this gap. We build MeDxBench, a large-scale benchmark of 4,421 clinical cases across 20 specialties. We further propose MeDxAgent, a multi-agent consultation system for interactive diagnosis, and systematically study its prompt-, flow- and agent-level design choices. MeDxAgent achieves a 10.3% accuracy gain over the baseline on MeDxBench, closing 52.3% of the gap to a full-information oracle. We find that specific design choices: collecting demographics first, passing summarized dialogue for diagnosis, and feeding candidate diagnoses for targeted questioning, improve accuracy, mirroring how physicians reason, though their effect emerges fully only in combination. Code and dataset will be released upon publication.

Akshat Sanghvi, Naren Akash, Raza Imam +2

Microsoft Research India · Institute of AI and Innovation, AIG Hospitals · Mohamed bin Zayed University of Artificial Intelligence

Multi-Agent LLM Systems Medical Diagnosis AI Agent Benchmarks Clinical Reasoning Medical AI Agents

Ask PDF

Mar 4, 2026

MIND: Unified Inquiry and Diagnosis RL with Criteria Grounded Clinical Supports for Psychiatric Consultation

Psychiatric consultation requires agents to elicit discriminative evidence, map uncertain narratives to diagnostic criteria, and decide when evidence suffices. Existing dialogue and retrieval-augmented systems condition policies on raw histories or appended passages, leaving observed evidence, missing checks, differentials, and support reliability entangled. We introduce MIND, a criteria-grounded evidence-state decision interface. At each turn, MIND constructs a typed state containing observed evidence, unresolved criterion checks, active differentials, criterion-linked supports, and reliability metadata. A training-split Psychiatric Reasoning Bank supplies gated supports, turning retrieval into state construction rather than prompt injection. The same state conditions action selection, process rewards, information-gain scoring, and trajectory rectification. Under EMR-grounded simulator protocols, MIND improves accuracy by 8.8 and 7.3 points over strong inference-only, RAG, and RL baselines, with gains transferring to public MDD-5k dialogues. Matched interventions attribute these gains to the shared state rather than prompt length, retrieval text, or formatting. MIND targets screening-level decision support, not autonomous diagnosis. Code is available at https://github.com/Lingxi-mental-health/MIND.Å

Guoyi Li, Shihao Xu, Jiatong Ma +4

EverMind AI Inc., California, USA · Venus Team, Ant Group · Tianqiao and Chrissy Chen Institute, Shanghai, China +3

Healthcare Evidence-Grounded Reasoning Medical Diagnosis Clinical Decision Support Mental Health RL for Language Model Reasoning+2

Ask PDF

Apr 7, 2026

SymptomWise: A Deterministic Reasoning Layer for Reliable and Efficient AI Systems

AI-driven symptom analysis systems face persistent challenges in reliability, interpretability, and hallucination. End-to-end generative approaches often lack traceability and may produce unsupported or inconsistent diagnostic outputs in safety-critical settings. We present SymptomWise, a framework that separates language understanding from diagnostic reasoning. The system combines expert-curated medical knowledge, deterministic codex-driven inference, and constrained use of large language models. Free-text input is mapped to validated symptom representations, then evaluated by a deterministic reasoning module operating over a finite hypothesis space to produce a ranked differential diagnosis. Language models are used only for symptom extraction and optional explanation, not for diagnostic inference. This architecture improves traceability, reduces unsupported conclusions, and enables modular evaluation of system components. Preliminary evaluation on 42 expert-authored challenging pediatric neurology cases shows meaningful overlap with clinician consensus, with the correct diagnosis appearing in the top five differentials in 88% of cases. Beyond medicine, the framework generalizes to other abductive reasoning domains and may serve as a deterministic structuring and routing layer for foundation models, improving precision and potentially reducing unnecessary computational overhead in bounded tasks.

Isaac Henry, Avery Byrne, Christopher Giza +2

University of California Los Angeles, David Geffen School of Medicine · University of Southern California

Clinical Decision Support Clinical Reasoning Differential Diagnosis Neuro-Symbolic Reasoning

Ask PDF

Abstract

Medical Diagnosis

Explore similar work

MeDxAgent: Multi-Agent Consultation for Interactive Medical Diagnosis

MIND: Unified Inquiry and Diagnosis RL with Criteria Grounded Clinical Supports for Psychiatric Consultation

SymptomWise: A Deterministic Reasoning Layer for Reliable and Efficient AI Systems