A Training-Free Symbolic Baseline for Interactive Medical Diagnosis: Automatic Rule-Slot Derivation and Weighted Evidence Combination
DOI:
https://doi.org/10.62411/jcta.17677Keywords:
Expected information gain, Explainable clinical decision support, Medical automatic diagnosis, Naive Bayes, Reinforcement learning, Rule-based expert system, Symptom inquiry, Training-free diagnosisAbstract
Interactive medical diagnosis requires sequential symptom elicitation before a diagnosis can be established. Although recent approaches employ reinforcement learning and transformer-based agents, their performance has not been adequately contextualized against a classical probabilistic baseline. This study introduces a training-free symbolic baseline that derives conditional symptom probabilities and three rule slots—must-have, exclusion, and critical—from co-occurrence counts in simulated patient records. Candidate diseases are ranked using observed evidence, while subsequent questions are selected by expected information gain. Evaluation on five benchmarks from a published reinforcement learning study reveals distinct performance regimes. On MedlinePlus (893 diseases), the proposed method achieves 99.27 ± 0.09% accuracy with an average of 1.75 questions, outperforming the published baseline (96.2 ± 0.1%; 3.41 questions). Across four SymCat variants, it improves accuracy by 3.03–7.70 percentage points but requires approximately twice as many questions; at a matched budget of 8.98 questions, the reference method achieves higher accuracy, highlighting a trade-off between diagnostic accuracy and interaction cost. Ablation experiments across six datasets show that removing weighted evidence combination reduces accuracy by 35.4–65.4 percentage points, while eliminating symptom inquiry reduces it by 23.3–66.8 percentage points. The safety veto removes up to 50.2% of candidates on average but changes the final diagnosis in at most 0.17% of consultations, indicating limited incremental impact on predictive accuracy under the evaluated configurations. Rule-base derivation takes 1.8 s for 893 diseases on a CPU. As all benchmarks rely on simulated patients, these findings establish benchmark-level performance rather than clinical effectiveness.References
C. Yu, J. Liu, S. Nemati, and G. Yin, “Reinforcement Learning in Healthcare: A Survey,” ACM Comput. Surv., vol. 55, no. 1, pp. 1–36, Jan. 2023, doi: 10.1145/3477600.
Z. Wei et al., “Task-oriented Dialogue System for Automatic Diagnosis,” in Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers), 2018, pp. 201–207. doi: 10.18653/v1/P18-2033.
Y.-S. Peng, K.-F. Tang, H.-T. Lin, and E. Chang, “REFUEL: exploring sparse features in deep reinforcement learning for fast disease diagnosis,” in Advances in Neural Information Processing Systems 31 (NeurIPS 2018), 2018, vol. 31, pp. 7322–7331. [Online]. Available: https://proceedings.neurips.cc/paper_files/paper/2018/hash/b5a1d925221b37e2e399f7b319038ba0-Abstract.html
H.-C. Kao, K.-F. Tang, and E. Chang, “Context-Aware Symptom Checking for Disease Diagnosis Using Hierarchical Reinforcement Learning,” Proc. AAAI Conf. Artif. Intell., vol. 32, no. 1, Apr. 2018, doi: 10.1609/aaai.v32i1.11902.
C. Zhong et al., “Hierarchical reinforcement learning for automatic disease diagnosis,” Bioinformatics, vol. 38, no. 16, pp. 3995–4001, Aug. 2022, doi: 10.1093/bioinformatics/btac408.
L. Xu, Q. Zhou, K. Gong, X. Liang, J. Tang, and L. Lin, “End-to-End Knowledge-Routed Relational Dialogue System for Automatic Diagnosis,” Proc. AAAI Conf. Artif. Intell., vol. 33, no. 01, pp. 7346–7353, Jul. 2019, doi: 10.1609/aaai.v33i01.33017346.
Y. Xia, J. Zhou, Z. Shi, C. Lu, and H. Huang, “Generative Adversarial Regularized Mutual Information Policy Gradient Framework for Automatic Diagnosis,” Proc. AAAI Conf. Artif. Intell., vol. 34, no. 01, pp. 1062–1069, Apr. 2020, doi: 10.1609/aaai.v34i01.5456.
J. Lin, L. Xu, Z. Chen, and L. Lin, “Towards a reliable and robust dialogue system for medical automatic diagnosis,” OpenReview.net. 2020. [Online]. Available: https://openreview.net/forum?id=TCAmP8zKZ6k
H. Yuan and S. Yu, “Efficient symptom inquiring and diagnosis via adaptive alignment of reinforcement learning and classification,” Artif. Intell. Med., vol. 148, p. 102748, Feb. 2024, doi: 10.1016/j.artmed.2023.102748.
W. Chen, C. Zhong, J. Peng, and Z. Wei, “DxFormer: a decoupled automatic diagnostic system based on decoder–encoder transformer with dense symptom representations,” Bioinformatics, vol. 39, no. 1, p. btac744, Jan. 2023, doi: 10.1093/bioinformatics/btac744.
F. Zhang, G. Sang, Z. Liu, H. Lin, and Y. Zhang, “A doctor’s diagnosis experience enhanced transformer model for automatic diagnosis,” Eng. Appl. Artif. Intell., vol. 134, p. 108675, Aug. 2024, doi: 10.1016/j.engappai.2024.108675.
C. Combi et al., “A manifesto on explainability for artificial intelligence in medicine,” Artif. Intell. Med., vol. 133, p. 102423, Nov. 2022, doi: 10.1016/j.artmed.2022.102423.
K. Zhang, D. Wang, F. Lin, J. Xie, and W. Zhou, “A comprehensive review of explainable artificial intelligence in healthcare methods, evaluation, and clinical integration,” iScience, vol. 29, no. 3, p. 115026, Mar. 2026, doi: 10.1016/j.isci.2026.115026.
S. Roychowdhury, V. Lanfranchi, and S. Mazumdar, “Evaluating explanation performance for clinical decision support systems for non-imaging data: A systematic literature review,” Comput. Biol. Med., vol. 197, p. 110944, Oct. 2025, doi: 10.1016/j.compbiomed.2025.110944.
B. M. Gebremariam, G. T. Aboye, A. A. Dessalegn, and G. L. Simegn, “Rule-based expert system for the diagnosis of maternal complications during pregnancy: For low resource settings,” Digit. Heal., vol. 10, Jan. 2024, doi: 10.1177/20552076241230073.
M.-E. Vidal, Y. Chudasama, H. Huang, D. Purohit, and M. Torrente, “Integrating Knowledge Graphs with Symbolic AI: The Path to Interpretable Hybrid AI Systems in Medicine,” J. Web Semant., vol. 84, p. 100856, Jan. 2025, doi: 10.1016/j.websem.2024.100856.
Y. Deng, S. Zhao, Y. Miao, J. Zhu, and J. Li, “MedKA: A knowledge graph-augmented approach to improve factuality in medical Large Language Models,” J. Biomed. Inform., vol. 168, p. 104871, Aug. 2025, doi: 10.1016/j.jbi.2025.104871.
W. Wallace et al., “The diagnostic and triage accuracy of digital and online symptom checker tools: a systematic review,” npj Digit. Med., vol. 5, no. 1, p. 118, Aug. 2022, doi: 10.1038/s41746-022-00667-w.
E. Riboli-Sasco et al., “Triage and Diagnostic Accuracy of Online Symptom Checkers: Systematic Review,” J. Med. Internet Res., vol. 25, p. e43803, Jun. 2023, doi: 10.2196/43803.
R. S. Ledley and L. B. Lusted, “Reasoning Foundations of Medical Diagnosis,” Science (80-. )., vol. 130, no. 3366, pp. 9–21, Jul. 1959, doi: 10.1126/science.130.3366.9.
G. A. Gorry and G. O. Barnett, “Experience with a model of sequential diagnosis,” Comput. Biomed. Res., vol. 1, no. 5, pp. 490–507, May 1968, doi: 10.1016/0010-4809(68)90016-5.
F. T. de Dombal, D. J. Leaper, J. R. Staniland, A. P. McCann, and J. C. Horrocks, “Computer-aided Diagnosis of Acute Abdominal Pain,” BMJ, vol. 2, no. 5804, pp. 9–13, Apr. 1972, doi: 10.1136/bmj.2.5804.9.
R. A. Miller, H. E. Pople, and J. D. Myers, “Internist-I , an Experimental Computer-Based Diagnostic Consultant for General Internal Medicine,” N. Engl. J. Med., vol. 307, no. 8, pp. 468–476, Aug. 1982, doi: 10.1056/NEJM198208193070803.
M. A. Shwe et al., “Probabilistic Diagnosis Using a Reformulation of the INTERNIST-1/QMR Knowledge Base,” Methods Inf. Med., vol. 30, no. 04, pp. 241–255, Feb. 1991, doi: 10.1055/s-0038-1634846.
D. E. Heckerman, E. J. Horvitz, and B. N. Nathwani, “Toward Normative Expert Systems: Part I The Pathfinder Project,” Methods Inf. Med., vol. 31, no. 02, pp. 90–105, Feb. 1992, doi: 10.1055/s-0038-1634867.
D. V. Lindley, “On a Measure of the Information Provided by an Experiment,” Ann. Math. Stat., vol. 27, no. 4, pp. 986–1005, Dec. 1956, doi: 10.1214/aoms/1177728069.
R. Howard, “Information Value Theory,” IEEE Trans. Syst. Sci. Cybern., vol. 2, no. 1, pp. 22–26, 1966, doi: 10.1109/TSSC.1966.300074.
K. Musgrave, S. Belongie, and S.-N. Lim, “A Metric Learning Reality Check,” in Proceedings of the European Conference on Computer Vision, 2020, pp. 681–699. doi: 10.1007/978-3-030-58595-2_41.
M. Ferrari Dacrema, P. Cremonesi, and D. Jannach, “Are we really making much progress? A worrying analysis of recent neural recommendation approaches,” in Proceedings of the 13th ACM Conference on Recommender Systems, Sep. 2019, pp. 101–109. doi: 10.1145/3298689.3347058.
P. Henderson, R. Islam, P. Bachman, J. Pineau, D. Precup, and D. Meger, “Deep Reinforcement Learning That Matters,” Proc. AAAI Conf. Artif. Intell., vol. 32, no. 1, Apr. 2018, doi: 10.1609/aaai.v32i1.11694.
W. He, X. Mao, C. Ma, Y. Huang, J. M. Hernàndez-Lobato, and T. Chen, “BSODA: A Bipartite Scalable Framework for Online Disease Diagnosis,” in Proceedings of the ACM Web Conference 2022, Apr. 2022, pp. 2511–2521. doi: 10.1145/3485447.3512123.
M. Werthaim, M. Kimhi, A. Apartsin, and Y. Aperstein, “A benchmark for evaluating diagnostic questioning efficiency of LLMs in patient conversations,” Sci. Rep., vol. 16, no. 1, p. 6121, Jan. 2026, doi: 10.1038/s41598-026-37022-y.
H. Nori et al., “Sequential Diagnosis with Language Models,” arXiv preprint. Jul. 02, 2025. [Online]. Available: http://arxiv.org/abs/2506.22405
G. A. Prenosil et al., “Neuro-symbolic AI for auditable cognitive information extraction from medical reports,” Commun. Med., vol. 5, no. 1, p. 491, Nov. 2025, doi: 10.1038/s43856-025-01194-x.
M. Y. H. Wong, J. C. L. Ong, N. Shah, D. C. Klonoff, and B. Sheng, “Neuro-symbolic artificial intelligence in medicine,” Nat. Biomed. Eng., vol. 10, no. 8, pp. 1507–1512, Jul. 2026, doi: 10.1038/s41551-026-01728-1.
P. Yang et al., “LMKG: A large-scale and multi-source medical knowledge graph for intelligent medicine applications,” Knowledge-Based Syst., vol. 284, p. 111323, Jan. 2024, doi: 10.1016/j.knosys.2023.111323.
X. Li and Q. Shen, “A hybrid framework based on knowledge distillation for explainable disease diagnosis,” Expert Syst. Appl., vol. 238, p. 121844, Mar. 2024, doi: 10.1016/j.eswa.2023.121844.
A. F. Tchango, R. Goel, Z. Wen, J. Martel, and J. Ghosn, “DDXPlus: A New Dataset For Automatic Medical Diagnosis,” in NIPS’22: Proceedings of the 36th International Conference on Neural Information Processing Systems, Oct. 2022, pp. 31306–31318. doi: 10.48550/arXiv.2205.09148.
E. Winarno, I. M. Nur, A. Karim, S. Amri, I. E. Wirdati, and P. W. Adi, “An Explainable Multimodal Framework for Chest X-Ray Alert Classification Using Radiology Reports and Images,” J. Comput. Theor. Appl., vol. 3, no. 4, pp. 647–666, May 2026, doi: 10.62411/jcta.16023.
A. S. Egbunu and A. M. Okedoye, “Harnessing Artificial Intelligence for Early Disease Detection: Opportunities and Challenges in Modern Healthcare,” J. Comput. Theor. Appl., vol. 3, no. 3, pp. 384–401, Feb. 2026, doi: 10.62411/jcta.15367.
Downloads
Published
How to Cite
Issue
Section
License
Copyright (c) 2026 Naufal Denta Elsatya Nugraha, Dhendra Marutho, Muhammad Sam’an, Wendy Sarasjati, De Rosal Ignatius Moses Setiadi

This work is licensed under a Creative Commons Attribution 4.0 International License.














