AI Systematic Review & Abstract Screening Screen 5,000 papers in an afternoon with AI

Lumina's AI learns your inclusion criteria and ranks papers by relevance. Screen the top 20% to find 95% of relevant studies.

Plays nicely with

PubMed Semantic Scholar OpenAlex EndNote Zotero
AI Systematic Review Software - Lumina Abstract Screening Interface

Relevance Score

98% Match

What is Lumina Systematic Review Software?

Lumina is a web application for researchers conducting systematic literature reviews. Instead of manually screening thousands of papers (which can take 40+ hours), Lumina's AI learns your inclusion criteria and ranks papers by relevance—so you focus on what matters.

😫

Without Lumina

  • Screening 5,000 titles manually
  • Takes 40+ hours of tedious reading
  • High risk of missing relevant studies
  • Reviewing unorganized, non-prioritized lists
🚀

With Lumina The Better Way

  • AI automatically ranks papers by relevance
  • Screen just the top 20% to find 95% of hits
  • Finish screening in hours, not weeks
  • Get a second opinion from Pixel-bot AI
See It In Action

Systematic Review Features, Beautifully Simple

Every feature is designed to save you time without sacrificing rigor.

Literature Alerts

3 active alerts

BPA endocrine disruption Active
Weekly PubMed, OpenAlex 12 new papers
Machine learning diagnostics Active
Daily Semantic Scholar 5 new papers
Stay Updated

Literature Alerts

Never miss a relevant paper again. Set up automated alerts for your search queries and get notified when new papers are published across your favourite databases. New papers are automatically imported into your project.

  • Daily, weekly, or monthly email notifications
  • Auto-import new papers into your project
  • Monitor multiple databases simultaneously
Core Feature

AI-Powered Screening with Relevance Scores

Our unique dual-scoring system combines semantic embeddings with a machine learning classifier that learns from your decisions in real-time. Every paper gets a relevance score—so you always screen the most promising papers first.

  • Find 95% of relevant studies by screening just 20%
  • Keyboard shortcuts (I/E/M) for blazing-fast decisions
  • AI explains why each paper may be relevant
AI Screening Interface with Relevance Scores
Search 6 Academic Databases for Systematic Reviews
Discovery

Search 6 Academic Databases

No more exporting and importing. Search PubMed, Semantic Scholar, OpenAlex, CrossRef, and more—right from Lumina. Import results with one click.

  • PubMed, Semantic Scholar, OpenAlex, CrossRef
  • One-click import to your project
  • Or upload database export files

Everything You Need for a Thorough Systematic Review

Advanced features built-in to save you time and ensure compliance.

Pixel-Bot AI Assistant

Get a second opinion from AI when unsure. Pixel-Bot explains exactly why a paper may be relevant to your unique criteria.

🤝

Team Collaboration & Dual Screening

Blinded dual-review mode with automated conflict detection and a seamless resolution workflow. Minimizes bias.

📋

Automatic Deduplication

Fuzzy matching automatically detects near-duplicates across multiple databases. Review them side-by-side before merging.

🎯

Data-Driven Stopping Rules

Know exactly when you've screened enough. Our stopping strategies save you hours of time without sacrificing recall.

📊

Auto-Generated PRISMA Diagram

Ready to publish? Lumina automatically creates a PRISMA 2020-compliant flow diagram based on your screening data.

📝

AI Transparency Report

Publish with confidence. Automatically generate PRISMA-trAIce and RAISE compliant methods paragraphs and audit logs.

Learn more
Swipe to explore →

Accurate Machine Learning Relevance Scores

See exactly why Lumina recommends a paper. It doesn't just score; it explains.

99% Match Example Abstract: AI Diagnostics

Deep learning model for early detection of Alzheimer's disease using MRI scans

Background: Early diagnosis of Alzheimer's disease (AD) is crucial for effective intervention. Objective: We propose a deep learning framework to detect early-stage AD from structural MRI data.

Methods: We trained a 3D convolutional neural network (CNN) on a dataset of 2,400 MRI scans from the ADNI database. The model architecture uses attention mechanisms to focus on hippocampal atrophy patterns.

Results: The model achieved an accuracy of 96.4% and an AUC of 0.98 in distinguishing mild cognitive impairment (MCI) from cognitively normal subjects.

Conclusion: This automated approach shows promise as a supportive diagnostic tool in clinical settings.

Lumina AI Analysis

Methodology Match

Identified use of "convolutional neural network (CNN)" and "MRI scans", matching your inclusion criteria for deep learning methodologies in neuroimaging.

Outcome Match

Detected high performance metrics ("accuracy of 96.4%", "AUC of 0.98") for "mild cognitive impairment", highly relevant to your research focus.

How AI Abstract Screening Works

Three simple steps to accelerate your review

1

Import

Search PubMed, OpenAlex, and 4+ databases directly—or upload export files from major scientific databases.

How to export from PubMed, Scopus, Web of Science, and more

2

Train

Make a few screening decisions. The AI learns your criteria in real-time, getting smarter with each choice.

3

Accelerate

Papers are ranked by relevance. Screen the top 20% to find 95% of relevant studies—stop when you're confident.

Why Choose Lumina for your Literature Reviews?

Tailored for every stage of your research journey.

🎓

PhD Students

Finish screening 5,000+ papers in days, not months. Export PRISMA diagrams instantly.

99%
Include
👥

Research Teams

Blinded dual-review mode and automatic conflict resolution for rigorous, unbiased screening.

Conflict Detected Action Required
Reviewer A Included
Reviewer B Excluded
🩺

Clinicians

AI highlights key findings so you can screen during coffee breaks. Mobile-friendly.

"We were dreading our systematic review until we found Lumina. It saved me and my team weeks of painfully tedious manual screening. The AI-powered relevance scoring is incredibly accurate and allowed me to focus purely on the literature that mattered."

S

Susan M.

Social Science Researcher

Free Tool

Need a PRISMA Flow Diagram?

Generate an official PRISMA 2020 flow diagram for your systematic review in seconds. No signup required.

Create Diagram Free

Frequently Asked Questions

Is there a free trial?

Yes! We offer a 14-day money-back guarantee on all plans. You can also try our public interactive demo without creating an account.

Can I collaborate with my team?

Absolutely. The Pro plan includes unlimited team members, real-time collaboration, and dual-review mode to ensure screening accuracy.

Is my data secure?

Yes. We are GDPR compliant and host all data within the EU (Germany). Your research data is private and never shared with third parties.

What format can I export to?

You can export your results to CSV, Excel, or RIS formats at any time. Compatible with EndNote, Zotero, Covidence, and Rayyan.

Ready to screen 10x faster?

Join hundreds of researchers accelerating their systematic reviews. Start your 14-day free trial today.

No credit card required for the 14-day free trial.