AI-Powered Differential Diagnosis with Explainable Recommendation for PHC Doctors
Authors
Department of CSE, Acharya Institute of Technology, Bangalore, India (India)
Department of CSE, Acharya Institute of Technology, Bangalore, India (India)
Department of CSE, Acharya Institute of Technology, Bangalore, India (India)
Department of CSE, Acharya Institute of Technology, Bangalore, India (India)
Article Information
DOI: 10.51583/IJLTEMAS.2026.151000039
Subject Category: Artificial Intelligence
Volume/Issue: 15/10 | Page No: 469-476
Publication Timeline
Submitted: 2026-10-15
Accepted: 2026-10-22
Published: 2026-11-03
Abstract
We built a Clinical Decision Support System (CDSS) for primary healthcare in India, following the National Health Mission (NHM) Standard Treatment Guidelines. It combines a React.js frontend, a FastAPI backend, and a machine learning inference engine to help clinicians reach a differential diagnosis and plan treatment. Natural language processing (NLP) can pull symptoms, vital signs, and laboratory results out of free-text clinical notes, and a Random Forest classifier then analyses them. We trained the classifier on synthetic data generated from the NHM guidelines, covering 38 disease categories that include respiratory, cardiovascular, neurological, vector-borne, gastrointestinal, and renal conditions. For each case the model returns the five most probable diagnoses with confidence scores. SHAP explanations show which inputs drove each prediction, a rule-based engine flags NHM danger signs, and a retrieval-augmented pipeline looks up the matching treatment protocol. On a held-out test set the classifier reached about 92% accuracy, and the full pipeline finished within the three-second target.
Keywords
Natural language processing (NLP), Retrieval-Augmented (RAG), Red Flag Detection, FastAPI, Primary Healthcare, Vector Database, Differential Diagnosis
This article is licensed under a Creative Commons Attribution 4.0 International License (CC BY 4.0).
Downloads
References
1. S. Kumari, R. Sharma, and A. Gupta, “Diagnostic Errors in Primary Healthcare Settings of Low- and Middle-Income Countries: A Systematic Review,” Journal of Global Health, vol. 14, 2024. [Google Scholar] [Crossref]
2. P. Rajpurkar, E. Chen, O. Banerjee, and E. Topol, “AI in Health and Medicine,” Nature Medicine, vol. 30, no. 3, pp. 1–10, 2024. [Google Scholar] [Crossref]
3. L. Yang, A. Shami, and T. Li, “Explainable AI-Driven Clinical Decision Support: A SHAP-Based Framework for Transparent Differential Diagnosis,” IEEE Access, vol. 12, pp. 45678–45692, 2024. [Google Scholar] [Crossref]
4. W. Huang, K. Zhao, and M. Chen, “Retrieval-Augmented Generation in Healthcare: A Comprehensive Review of Architectures and Clinical Applications,” MDPI Applied Sciences, vol. 15, no. 4, 2025. [Google Scholar] [Crossref]
5. J. Kang, S. Patel, and H. Lin, “Improving Large Language Model Applications in Biomedicine with Retrieval-Augmented Generation: A Systematic Review and Meta-Analysis,” Journal of the American Medical Informatics Association (JAMIA), vol. 32, no. 1, pp. 88–102, 2025. [Google Scholar] [Crossref]
6. R. Singh, N. Mehta, and P. Agarwal, “Natural Language Processing for Clinical Symptom Extraction: Methods, Challenges, and Applications in Low-Resource Healthcare Environments,” Computers in Biology and Medicine, vol. 168, 2024. [Google Scholar] [Crossref]
7. A. Verma, S. Kumar, and D. Rao, “Ensemble Machine Learning for Febrile Illness Differential Diagnosis in India: Alignment with National Health Mission Guidelines,” Journal of Biomedical Informatics, vol. 151, 2024. [Google Scholar] [Crossref]
8. Z. Lewis, M. Osei, and A. Park, “SHAP-Based Interpretability in Random Forest Models for Multi-Class Medical Diagnosis: A Comparative Study,” PLOS ONE, vol. 19, no. 6, 2024. [Google Scholar] [Crossref]
9. T. Nguyen, B. Raj, and F. Al-Turjman, “Synthetic Data Generation for Clinical Machine Learning: Balancing Fidelity, Privacy, and Model Performance,” IEEE Journal of Biomedical and Health Informatics, vol. 28, no. 4, pp. 2010–2022, 2025. [Google Scholar] [Crossref]
10. Ministry of Health and Family Welfare, Government of India, “Standard Treatment Guidelines — National Health Mission,” National Health Systems Resource Centre (NHSRC), New Delhi, 2024. [Google Scholar] [Crossref]
Metrics
Views & Downloads
Similar Articles
- Impact of AI Tools on Student Learning
- Repogent: An Autonomous Multi-Agent System for End-To-End Repository Maintenance
- Trust-Centric Federated Learning Using Blockchain-Enabled Smart Contracts: An Analysis of Data Privacy, Model Integrity, and Decentralized AI Governance
- Can AI Care? Rethinking Ethics, Clinical Integrity, and Professional Responsibility in the Digital Age of Mental Healthcare
- AI-Based Smart Irrigation System Using IoT and Machine Learning for Precision Agriculture
This work is licensed under CC BY 4.0