AI Co-Scientist for Knowledge Synthesis in Medical Contexts: A Proof of Concept
Arya Rahgozar, Pouria Mortezaagha
TL;DR
This paper tackles research waste in biomedical knowledge synthesis by introducing an AI co-scientist that formalizes PICOS (Population, Intervention, Comparator, Outcome, Study design) and integrates automated screening with retrieval-augmented, evidence-grounded synthesis. The system combines a relational data layer, a vector semantic store, and a Neo4j knowledge graph, orchestrated by an agent that enforces grounding and provenance. It features a domain-adapted multi-task transformer (Kernel) for PICOS and study-design classification, a graph-based retrieval pathway (Neo4j GraphRAG), a metadata-aware semantic retrieval path (pgVector), BERTopic for topic modeling, and a conversational recommender with a live database interface. Evaluations in dementia–sport and non-communicable diseases show high PICOS-detection accuracy (Bi-LSTM ≈87%, Kernel ≈95–96% for study-design) and effective RAG performance for constrained, cross-study queries, revealing redundancy and underexplored areas; overall, the architecture demonstrates scalable, explainable, and domain-agnostic potential to reduce research waste and accelerate evidence synthesis in biomedical domains.
Abstract
Research waste in biomedical science is driven by redundant studies, incomplete reporting, and the limited scalability of traditional evidence synthesis workflows. We present an AI co-scientist for scalable and transparent knowledge synthesis based on explicit formalization of Population, Intervention, Comparator, Outcome, and Study design (PICOS). The platform integrates relational storage, vector-based semantic retrieval, and a Neo4j knowledge graph. Evaluation was conducted on dementia-sport and non-communicable disease corpora. Automated PICOS compliance and study design classification from titles and abstracts were performed using a Bidirectional Long Short-Term Memory baseline and a transformer-based multi-task classifier fine-tuned from PubMedBERT. Full-text synthesis employed retrieval-augmented generation with hybrid vector and graph retrieval, while BERTopic was used to identify thematic structure, redundancy, and evidence gaps. The transformer model achieved 95.7% accuracy for study design classification with strong agreement against expert annotations, while the Bi-LSTM achieved 87% accuracy for PICOS compliance detection. Retrieval-augmented generation outperformed non-retrieval generation for queries requiring structured constraints, cross-study integration, and graph-based reasoning, whereas non-retrieval approaches remained competitive for high-level summaries. Topic modeling revealed substantial thematic redundancy and identified underexplored research areas. These results demonstrate that PICOS-aware and explainable natural language processing can improve the scalability, transparency, and efficiency of evidence synthesis. The proposed architecture is domain-agnostic and offers a practical framework for reducing research waste across biomedical disciplines.