KEY TAKEAWAYS
- The study aimed to identify primary topics in GB research and assess their trend dynamics using NLP.
- The results demonstrated NLP’s ability to reveal key insights in GB literature, guiding research and policy directions.
Mert Karabacak and the team spearheaded a study that aimed to explore the primary topics in glioblastoma (GB) research, assessing trends to identify “hot” or “cold” areas. They also aimed to highlight how NLP enhances research synthesis, crucial for dissecting the brain cancer literature landscape effectively.
Researchers queried the Scopus database using “glioblastoma” in the “TITLE” and “KEY” fields. They employed BERTopic, an NLP-based method for probabilistic topic modeling, setting a minimum topic size of 300 documents and a 5% probability cutoff for outlier detection.
Topics were labeled using keywords and representative documents, visualized through word clouds. Linear regression models were applied to discern trends in topic popularity across decades, distinguishing between “hot” and “cold” areas of research.
The results revealed that the topic modeling (TM) analysis classified 43,329 articles into 15 distinct topics. Predominant topics included Genomics, Survival, Drug Delivery, and Imaging, whereas less common topics included Surgical Resection, MGMT Methylation, and Exosomes.
During the 2020s, the most active research areas were Viruses and Oncolytic Therapy, Anticancer Compounds, and Exosomes, contrasting with less explored topics such as Surgical Resection, Angiogenesis, and Tumor Metabolism.
The study concluded that the NLP methodology provided a comprehensive analysis of GB literature, uncovering nuanced historical and contemporary patterns beyond traditional techniques.
These findings provided valuable insights for shaping research agendas, informing policy decisions, and identifying emerging trends. The approach demonstrated potential applicability across diverse research fields for summarizing and exploring scholarly literature, facilitating future investigations.
No funding or grants were provided for the study.
Source: https://link.springer.com/article/10.1007/s11060-024-04762-8
Karabacak M, Jagtiani P, Carrasquilla A, et al. (2024). “Simplifying synthesis of the expanding glioblastoma literature: a topic modeling approach.” J Neurooncol (2024). https://doi.org/10.1007/s11060-024-04762-8
