MARS (Medical Assistant for Research Synthesis) is an AI-powered assistant that helps you research topics in medicine and document them in structured, comprehensive reports. It automates the entire process from developing research strategies to critical appraisal, synthesis, and professional report generation making it ideal for medical researchers, clinicians, students, and anyone seeking evidence-based insights in medicine.
This is a python/next js application. Backend is python, Frontend is Nextjs. The backend uses crewai framework to assist researchers to search and document research topics. Frontend is a simple nextjs application with simple, cool features with firebase integration for authentication.
- Automated Research Strategy: Identifies the best databases (e.g., PubMed, MEDLINE) and search strategies for any medical topic.
- Literature Search & Curation: Conducts systematic searches, applies inclusion/exclusion criteria, and curates the most relevant articles.
- Critical Appraisal: Evaluates study design, methodology, bias, and evidence quality for top research articles.
- Synthesis & Trend Identification: Summarizes key findings, trends, controversies, and knowledge gaps in the field.
- Executive Summary: Produces concise summaries suitable for non-specialist audiences.
- Detailed Report Generation: Creates structured reports with introduction, methodology, results, discussion, and conclusions.
- Professional Formatting: Outputs clean, error-free Markdown (or PDF) reports with title pages, tables of contents, and references.
- Define a Topic: Specify the medical topic you want to research (e.g., "Post-operative issues and patient care").
- MARS Agents in Action:
- Develop a research strategy tailored to your topic.
- Search and curate the latest and most relevant literature.
- Critically appraise and synthesize the findings.
- Receive Outputs: Get a well-formatted, referenced document with all findings, trends etc.
git clone https://github.com/Atharv-web/MARS.git
cd MARS
cd med-researcher
npm installcd MARS
cd researcher
pip install -r requirements.txt- Configure Backend: run the python based backend
cd researcher/scr/researcher
uvicorn main:app --reload- Run frontend Script:
cd med-researcher
npm run dev-
Provide Topic Input: Provide your input.
-
Review Outputs: Find research report in the
researcher/src/results/directory.
Reports include:
- Title page & table of contents
- Executive summary
- Introduction, methodology, results, discussion, and conclusion
- Key trends, gaps, and controversies
- References in standard citation style