Prediction of transformative breakthroughs in biomedical research.

Area of Science:

• Bibliometrics
• Artificial Intelligence
• Network Science

Background:

• Scientific breakthroughs accelerate discovery but are hard to predict.
• Efficiently identifying breakthrough-prone research areas is crucial for investment and progress.

Purpose of the Study:

• To develop a method for predicting scientific breakthroughs.
• To identify a signature in co-citation networks indicative of future breakthroughs.

Main Methods:

• Utilized artificial intelligence and machine learning (AI/ML) to analyze co-citation networks.
• Identified a signature characterized by bursts of novel concept papers, influential publications, and low topical cohesion.
• Analyzed data across two distinct 20-year periods to confirm conserved kinetics.

Main Results:

• A common signature in co-citation networks predicts medical research breakthroughs.
• The signature predicts discoveries on average 5 years in advance, with some cases up to 12 years.
• The kinetics of breakthrough formation were found to be conserved over time.

Conclusions:

• The developed AI/ML approach can accurately predict topics likely to produce future transformative discoveries.
• This method enhances the efficiency of research investments by identifying high-potential areas.
• Understanding the scientific process through network analysis can improve research returns.