Mode
Text Size
Log in / Sign up

Google NotebookLM achieves 99.77% critical accuracy and 89.3% workflow time reduction in systematic review data extractionAI tools can speed up research on inflammatory bowel disease

AI-generated summary of the cited source, checked by automated accuracy review. How we work

Key Takeaway
Note that Google NotebookLM significantly reduces workflow time while maintaining high critical accuracy in systematic reviews.

This systematic review evaluates the performance of Google NotebookLM as an AI-assisted tool for data extraction and generating consensus statements within a multidisciplinary systematic review focusing on Inflammatory Bowel Disease (IBD) and obesity. The study analyzed 57 articles to compare AI-generated outputs against conventional extraction benchmarks.

The analysis found that Google NotebookLM achieved a cell-level accuracy of 91.17% and a critical accuracy of 99.77%. Furthermore, the use of the tool resulted in an 89.3% reduction in workflow time, decreasing from 165.1 person-hours to 17.7 person-hours. Of the AI-derived statements, 85.7% were retained in the final expert-finalized set.

A primary limitation noted by the authors is extractive incompleteness. Despite these findings, the tool's utility is framed as a method for improving research efficiency rather than a clinical intervention. The results suggest that while AI can significantly accelerate the synthesis of complex data regarding IBD and obesity, human oversight remains necessary to ensure completeness.

How this fits prior evidence

This methodological report addresses a gap in research methodology by evaluating AI-assisted tools for systematic reviews. While prior coverage has focused on biological pathways like B cell and humoral immunity in Inflammatory Bowel Disease or the role of lymphatic dysfunction in sarcopenic obesity, this study focuses on the technical feasibility of synthesizing such complex data efficiently using Google NotebookLM.

Researchers are trying to find better ways to manage the complex links between obesity and inflammatory bowel disease. Because these conditions involve many different factors, gathering and organizing all the necessary data is a massive task that usually takes experts hundreds of hours to complete.

To solve this, researchers tested an AI tool called NotebookLM to help extract information from 57 different articles. The results were impressive. The AI achieved over 91 percent accuracy at the cell level and nearly 100 percent critical accuracy when identifying key facts. This means it could reliably identify important data points without making major factual errors.

Most importantly, using this tool cut the time needed for the research workflow by about 89 percent compared to traditional methods. While the AI is not a replacement for human experts, it showed great promise as a way to help medical teams work faster and more efficiently. One small note: the tool can sometimes miss some details during the extraction process.

What this means for you:
AI tools can cut research time by nearly 90 percent while maintaining high accuracy in complex medical data.

Common questions

How accurate is this AI tool at finding medical facts?

The study found that the AI had a cell-level accuracy of 91.17 percent. More importantly, it achieved a critical accuracy of 99.77 percent, which means it was highly reliable at identifying key information without making major factual errors during the data extraction process.

How much time does using AI save for researchers?

Using the AI tool resulted in an 89.3 percent reduction in workflow time compared to traditional methods. Specifically, it took about 17.7 person-hours with the AI versus 165.1 person-hours using conventional extraction techniques.

Is this a new treatment for inflammatory bowel disease?

No, this is not a medical treatment or drug. It is a methodological study testing how well artificial intelligence can help researchers organize and summarize complex data about conditions like obesity and inflammatory bowel disease.

Study Details

Study typeSystematic review
EvidenceLevel 1
PublishedJun 2026
View Original Abstract ↓
Background: Large language models (LLMs) offer promise for systematic review data extraction, but performance in complex multidisciplinary domains and utility for clinical statement generation remain insufficiently described. Objectives: To evaluate Google NotebookLM for AI-assisted data extraction and RAND/UCLA consensus statement generation in a systematic review of IBD, obesity, and cardiometabolic comorbidities. Methods: Studies were organized into domain-specific notebooks; structured prompts generated standardized evidence tables. Two independent reviewers validated outputs against full-text articles using a four-category error classification. Cell-level accuracy and critical accuracy (cells free of major factual errors) were the primary metrics; workflow time was compared against a published conventional extraction benchmark. Concordance between AI-generated and expert-finalized statements was assessed. Results: Across 57 articles, 1,710 data cells were extracted; 151 (8.83%) were flagged, yielding 91.17% cell-level accuracy. Major factual errors occurred in only 4 cells (0.23%), for a critical accuracy of 99.77%. Most errors were minor omissions (59.6%) or incomplete extractions (30.5%); domain error rates ranged from 7.08% to 11.33%. The pipeline required 17.7 versus a projected 165.1 person-hours (89.3% reduction). PICO-structured prompting generated 70 candidate statements; 58 of 112 finalized panel statements (51.8%) were AI-derived, and 85.7% were retained in the finalized set. Conclusion: Google NotebookLM demonstrates feasibility as a primary extraction and synthesis tool in a multidisciplinary systematic review, with extractive incompleteness as the principal limitation and substantial time savings over conventional approaches. Its novel application to RAND/UCLA consensus statement generation extends AI-assisted evidence synthesis to clinical consensus generation workflow.
Free Newsletter

Clinical research that matters. Delivered to your inbox.

Join thousands of clinicians and researchers. No spam, unsubscribe anytime.