An AI-Driven Live Systematic Reviews in the Brain-Heart Interconnectome: Minimizing Research Waste and Advancing Evidence Synthesis
Arya Rahgozar, Pouria Mortezaagha, Jodi Edwards, Douglas Manuel, Jessie McGowen, Merrick Zwarenstein, Dean Fergusson, Andrea Tricco, Kelly Cobey, Margaret Sampson, Malcolm King, Dawn Richards, Alexandra Bodnaruc, David Moher
TL;DR
This paper tackles inefficiencies and research waste in Brain-Heart Interconnectome evidence synthesis by introducing an AI-driven living systematic review system. The framework combines automated PICOS compliance detection, semantic search with vector embeddings, graph-based querying in Neo4j, BERTopic for topic modeling, and Retrieval-Augmented Generation to answer complex, multi-source queries. It demonstrates high performance in PICO detection (87% accuracy) and study-design classification (precision 91.4%, recall 100%, accuracy 95.7%), with RAG-3.5 outperforming standalone GPT-4 on domain-specific tasks. The system provides real-time updates, interactive dashboards, and conversational AI, offering a scalable, adaptable template for rigorous evidence synthesis across biomedical fields.
Abstract
The Brain-Heart Interconnectome (BHI) combines neurology and cardiology but is hindered by inefficiencies in evidence synthesis, poor adherence to quality standards, and research waste. To address these challenges, we developed an AI-driven system to enhance systematic reviews in the BHI domain. The system integrates automated detection of Population, Intervention, Comparator, Outcome, and Study design (PICOS), semantic search using vector embeddings, graph-based querying, and topic modeling to identify redundancies and underexplored areas. Core components include a Bi-LSTM model achieving 87% accuracy for PICOS compliance, a study design classifier with 95.7% accuracy, and Retrieval-Augmented Generation (RAG) with GPT-3.5, which outperformed GPT-4 for graph-based and topic-driven queries. The system provides real-time updates, reducing research waste through a living database and offering an interactive interface with dashboards and conversational AI. While initially developed for BHI, the system's adaptable architecture enables its application across various biomedical fields, supporting rigorous evidence synthesis, efficient resource allocation, and informed clinical decision-making.