Skip to content

DRE Assistant

The DRE Assistant is your AI research co-pilot that understands your project context.

Overview

DRE Assistant can:

  • Answer questions about your data
  • Run analyses on request
  • Generate draft sections for your paper
  • Find relevant citations
  • Explain statistical concepts

Accessing the Assistant

Click the ✨ Assistant button in the sidebar to open the chat panel.

Keyboard Shortcut

Press Ctrl/Cmd + K anywhere in MedTWIN.

Context Menu

Right-click any statistic or table and select Ask Assistant.

Example Conversations

Asking About Data

You: What are the top mortality predictors in my data?

DRE: Based on RUN-00234, the top predictors of 30-day mortality are:

  1. Emergency surgery (OR 3.8, 95% CI 2.1-6.8, p<0.001)
  2. Age ≥75 years (OR 2.41, 95% CI 1.4-4.2, p=0.003)
  3. Diabetes mellitus (OR 2.31, 95% CI 1.2-4.4, p=0.008)

Source: RUN-00234, Logistic Regression

Generating Content

You: Write the results section for mortality analysis.

DRE: Here's a draft Results section:

Results

Of 847 patients enrolled, 30-day mortality occurred in 27 (3.2%). [✓ Verified from RUN-00234]

In multivariable logistic regression, five independent predictors of mortality were identified (Table 2). Emergency surgery demonstrated the strongest association with mortality (OR 3.8, 95% CI 2.1-6.8, p<0.001), followed by age ≥75 years (OR 2.41, 95% CI 1.4-4.2, p=0.003).

[Insert in Paper]

Finding Citations

You: Find papers about diabetes and cardiac surgery mortality.

DRE: I found 12 relevant papers:

  1. Nashef et al. (2024) - "EuroSCORE III: Updated Risk Model"
  2. European Heart Journal, 847 citations

  3. Discusses diabetes as independent predictor

  4. Smith et al. (2023) - "Diabetes and Cardiac Surgery Outcomes"

  5. JAMA Surgery, 312 citations
  6. Reports 2.1x mortality risk with diabetes

[Add to References]

Quick Actions

Common tasks available with one click:

Action Description
📊 Generate Table 1 Create baseline characteristics table
📈 Run Analysis Execute configured analysis
📝 Draft Abstract Generate abstract from results
📚 Find Citations Search relevant literature
✓ Verify Stats Check all statistics in paper

Context Awareness

DRE Assistant knows:

  • Your uploaded data (variables, values)
  • Your analysis results (all runs)
  • Your paper draft (current content)
  • Your citations (added references)

This context enables natural questions like:

  • "What's the mean age in my data?"
  • "How does my mortality rate compare to literature?"
  • "Which statistics haven't been verified?"

Traceable Responses

Every response from DRE includes:

  • Source reference: Where the information came from
  • Confidence level: How certain the answer is
  • Verification status: If statistics are verified

No Hallucinations

DRE only reports statistics from your actual data and analyses. If it doesn't know, it will say so.

Limitations

DRE Assistant cannot:

  • Access data from other projects
  • Make clinical recommendations
  • Guarantee publication acceptance
  • Replace statistical review

Tips for Best Results

Getting Better Answers

  1. Be specific: "What's the mortality in patients over 75?" vs "Tell me about mortality"
  2. Reference specific analyses: "Using RUN-00234, ..."
  3. Ask for sources: "Show me where that number comes from"
  4. Iterate: Follow up for clarification

Feedback

Help improve DRE Assistant:

  • 👍 Thumbs up helpful responses
  • 👎 Thumbs down unhelpful ones
  • Add comments explaining why

Next Steps