This research introduces a hybrid system that overcomes traditional LLM limitations in medical literature searches by integrating domain-specific knowledge structures.

• Combines the reasoning capabilities of LLMs with structured medical ontologies (MeSH and UMLS)
• Achieves significantly higher precision and recall compared to standard LLM approaches in identifying relevant medical literature
• Provides transparent reasoning chains that medical professionals can review and trust
• Reduces the time and expertise barriers for conducting comprehensive systematic reviews in healthcare

This innovation matters for medical research by making systematic reviews more accessible and reliable, potentially accelerating evidence-based medicine adoption while maintaining scientific rigor.

NeuroLit Navigator: A Neurosymbolic Approach to Scholarly Article Searches for Systematic Reviews