Reviewer-controlled screening

Systematic review screening software for titles and abstracts

Import or search citations, rank the queue against your eligibility criteria, and keep human decisions, reviewer conflicts, notes, and exports in one traceable workflow.

Systematic review title and abstract screening software with AI priority score and reviewer decisions

One screening workflow from citation to decision

AI organizes attention; reviewers decide what moves forward.

01

Bring in citations

Upload RIS, NBIB, BibTeX, CSV, or tagged text, or use the integrated database search.

02

Prioritize the queue

Rank titles and abstracts against the project question and eligibility criteria from the first session.

03

Make reviewer decisions

Record Include, Exclude, or Maybe while notes and supporting context remain attached.

04

Resolve and export

Compare independent decisions, resolve conflicts, and export structured screening records.

What screening software should support

Explicit eligibility criteria

Reviewers need the same operational definitions in view. Lumina stores the research question and inclusion and exclusion criteria with the project.

Human-controlled AI prioritization

A relevance score can change screening order. It should not silently make final eligibility decisions or hide unscreened records.

Blinded dual screening

Two reviewers can assess records independently before decisions are compared and disagreements enter a conflict queue.

Traceable exports

Decision history, reasons, notes, reviewer activity, and review counts remain available for reporting and audit.

What Lumina does not automate

Lumina does not replace protocol design, database-specific search validation, risk-of-bias assessment, methodological judgement, or the reviewer responsible for a final inclusion decision. AI recommendations remain advisory.

Review the AI transparency approach →

Screening resources

See the screening workflow before creating a project

Use the public demo to inspect ranked citations, evidence context, and reviewer decisions with sample data.

Open the free demo →