Researchers from Mount Sinai Health System in New York set out to get an idea of the “hot” and “cold” topics in spine research over time. Their work, “Natural language processing reveals research trends and topics in The Spine Journal over two decades: a topic modeling study,” appears in the March 2024 edition of The Spine Journal.
“The field of spine research is rapidly evolving, with new research topics continually emerging,” co-author Konstantinos Margetis M.D., Ph.D., director of Complex Spine Surgery, Department of Neurosurgery, Mount Sinai Hospital, told OTW.
“Analyzing topics and trends in the literature can provide insights into the shifting research landscape. This study aimed to elucidate prevalent and emerging research topics and trends within The Spine Journal using a natural language processing technique called topic modeling. This is almost impossible to do it manually with conventional methods because it is very time intensive. Therefore, we used natural language processing methodology.”
To review articles from The Spine Journal, the team used BERTopic, a topic modeling technique rooted in natural language processing. The team directed BERTopic to “read” 3,358 documents and vectorize the content into 30 distinct topics.
The most frequently identified topics were:
- “Outcome Measures,”
- “Scoliosis,” and
- “Intradural Lesions.”
Throughout the history of the journal, the three hottest topics were:
- “Degenerative Cervical Myelopathy,”
- “Osteoporosis,” and
- “Opioid Use.”
Conversely, the coldest topics were:
- “Intradural Lesions,”
- “Extradural Tumors,” and
- “Vertebral Augmentation.”
Within the current decade, the hottest topics were:
- “Screw Biomechanics,”
- “Paraspinal Muscles,” and
- “Biologics for Fusion,”
And the cold topics were:
- “Intraoperative Blood Loss,”
- “Construct Biomechanics,” and
- “Material Science.”
Dr. Margetis added: “Rising trends, such as ‘Screw Biomechanics,’ ‘Paraspinal Muscles,’ and ‘Biologics for Fusion,’ denote rising themes of interest in spine research, stimulated by technological progress, novel findings, and evolving healthcare needs.”
“Declining trends, like ‘Intraoperative Blood Loss,’ ‘Construct Biomechanics,’ and ‘Material Science,’ suggest either maturing or underrepresented areas that might require further exploration.”
Looking forward, Dr. Margetis told OTW, “While the theoretical advantages of BERTopic over traditional methods like Latent Dirichlet Allocation are promising, comprehensive validation of these advantages remains challenging.”
“The gold standard for validation would involve human experts manually reviewing and categorizing thousands of abstracts—a task that is not just time-consuming but also practically unfeasible due to the sheer volume and complexity of the data. We acknowledge this as a significant limitation, and while we are confident in the potential of BERTopic based on its foundational methodologies and our observations, we recognize the need for rigorous validation.”
