> ## Documentation Index
> Fetch the complete documentation index at: https://docs-evisyn.grabmywork.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Full-Text Screening

> Retrieve PDFs automatically and run AI analysis on complete papers

After human review, the next steps are retrieving full-text PDFs and running a deeper AI analysis on the complete papers.

## Step 1: Full Paper Download

Navigate to **Full Paper Download** from your project dashboard.

<img src="https://mintcdn.com/ravian/YwBZFdlBbtQ_-NDF/images/full-paper-download.png?fit=max&auto=format&n=YwBZFdlBbtQ_-NDF&q=85&s=73346fbc16f2d69e40a5eba7df1ff2a3" alt="Full paper download page" width="2496" height="1758" data-path="images/full-paper-download.png" />

### Upload PDFs

Two upload zones are available:

<CardGroup cols={2}>
  <Card title="Upload PDFs for Existing Papers" icon="file-pdf">
    Drag and drop PDFs for papers already in your shortlist. EviSyn auto-matches PDFs to papers by **DOI**, **title**, or **filename**.
  </Card>

  <Card title="Upload Additional Papers" icon="file-circle-plus">
    Upload PDFs for papers from other sources (e.g., hand search, Scopus, CINAHL). These are added as new finalized papers with metadata auto-extracted from the PDF.
  </Card>
</CardGroup>

### Automatic Retrieval

EviSyn also attempts to retrieve PDFs automatically from:

| Source                   | Description                                                 |
| ------------------------ | ----------------------------------------------------------- |
| **PubMed Central (PMC)** | Free full-text for open-access articles                     |
| **Unpaywall**            | Discovers legally available green/gold open-access versions |
| **Semantic Scholar**     | S2 API for additional PDF lookup                            |

## Step 2: Full-Text Screening

Navigate to **Full-Text Screening** — the AI reads each complete paper and evaluates relevance.

<img src="https://mintcdn.com/ravian/YwBZFdlBbtQ_-NDF/images/fulltext-screening-2.png?fit=max&auto=format&n=YwBZFdlBbtQ_-NDF&q=85&s=a852bc5ed33de905a23f416f2c5f4291" alt="Full-text screening setup" width="3840" height="1760" data-path="images/fulltext-screening-2.png" />

### Configuration

Like TiAb screening, you can customize:

* **Inclusion Criteria** — defaults inherited from TiAb criteria, editable for deeper analysis
* **Exclusion Criteria** — specific to full-text evaluation
* **AI Confidence Threshold** — papers below this score are flagged as "Doubt"

### Stats Dashboard

| Metric               | Meaning                                               |
| -------------------- | ----------------------------------------------------- |
| **Total**            | Finalized papers eligible for full-text screening     |
| **Screened**         | Papers the AI has analyzed                            |
| **Included** (green) | Papers confirmed as relevant after full-text analysis |
| **Doubt** (orange)   | Uncertain — needs human review                        |
| **Excluded** (red)   | Not meeting criteria after full-text review           |
| **Pending**          | Awaiting screening                                    |
| **No Full Text**     | PDFs not yet available                                |

### Screening Controls

* **Screen All Pending** — Run AI screening on all papers with available full text
* **Select Not Started / Doubt / Excluded** — Filter papers by status
* **Screening / Retry Selected** — Re-screen specific papers

<img src="https://mintcdn.com/ravian/YwBZFdlBbtQ_-NDF/images/fulltext-papers-list.png?fit=max&auto=format&n=YwBZFdlBbtQ_-NDF&q=85&s=bbc42af90e7590d2a7d1d1c2a1d4b4c3" alt="Full-text screening paper list" width="3840" height="1766" data-path="images/fulltext-papers-list.png" />

### What the AI Analyzes

For each paper with a PDF, the AI:

* Reads the **entire document** (not just the abstract)
* Evaluates relevance **section by section** (Methods, Results, Population, Intervention, Outcomes)
* Generates a **relevance score** (0-100)
* Provides **section-by-section notes**
* Cites specific passages as evidence

## What's Next?

After full-text screening, proceed to [Data Extraction](/user-guide/data-extraction) to pull structured data from your included papers.