🔬MetaLens AI

How MetaLens AI Works

From keyword to comprehensive meta-analysis in seconds. Here is a detailed look at every step of the process.

⌨️
Step 1

Enter Your Medical Keywords

Start by typing a medical topic, drug name, disease, or clinical question into the search bar on the MetaLens AI homepage. You can use simple terms like "metformin diabetes" or more specific queries like "SGLT2 inhibitors heart failure mortality reduction." The more specific your query, the more focused and relevant your results will be.

Pro Tip:Use specific medical terminology rather than general language. For example, instead of "blood pressure medicine," try "ACE inhibitors hypertension outcomes." You can also include study type keywords like "randomized controlled trial" or "systematic review" to narrow results to particular study designs.
🔍
Step 2

PubMed Database Search

Behind the scenes, MetaLens AI connects to the National Library of Medicine's PubMed database through the official E-Utilities API. Your query is translated into an optimized PubMed search that scans over 40 million biomedical citations and abstracts. The system retrieves the most relevant and recent research papers, prioritizing high-quality studies such as randomized controlled trials, systematic reviews, and meta-analyses.

Depending on your subscription tier, MetaLens AI retrieves between 10 abstracts (Free tier) and 80 abstracts (Ultra tier) per query. Each abstract is fetched in real time, ensuring you always receive the most up-to-date research available in the PubMed database.

🤖
Step 3

AI-Powered Synthesis and Analysis

Once the relevant abstracts are retrieved, they are fed into Google's Gemini large language model, which has been specifically configured to perform structured meta-analysis. The AI reads and comprehends each abstract, identifying key findings, study methodologies, sample sizes, and conclusions. It then cross-references these findings across all retrieved papers to identify patterns of agreement, conflicting results, and knowledge gaps.

The synthesis process takes approximately 10 to 30 seconds, depending on the complexity of the topic and the number of abstracts being analyzed. During this time, the AI is performing work that would typically take a human researcher several hours to complete manually.

📊
Step 4

Review Your Structured Results

Your results are presented in a clear, professionally structured format that mirrors the conventions of academic systematic reviews. The report typically includes an executive summary of the overall findings, a breakdown of key themes and outcomes, areas where the literature shows strong consensus, any conflicting or contradictory results, clinical implications, and suggestions for further research.

Every claim in the report is backed by specific PubMed citations with clickable PMID links, so you can immediately verify any finding against the original source paper. You can also save your analysis results to your account for future reference, export them, or run additional queries to explore related topics.

💡 Tips for Getting the Best Results

1.

Be specific with your queries.Instead of searching for a broad term like "cancer," narrow it down to "immunotherapy non-small cell lung cancer survival" for more focused and relevant results.

2.

Use medical terminology. PubMed indexes papers using MeSH (Medical Subject Headings) terms. Using proper medical vocabulary will improve the quality of retrieved studies.

3.

Combine related terms.A query like "statin cardiovascular prevention elderly" will retrieve more targeted results than searching for each term separately.

4.

Check the cited sources. Always click through to the original PubMed articles to verify key claims, especially if you plan to use the information in clinical or academic contexts.

5.

Upgrade for deeper analysis. If you need more comprehensive coverage, the Pro and Ultra tiers analyze significantly more abstracts per query (up to 80), providing a more thorough synthesis of the available evidence.