Meet MARK, your dedicated advisor for healthcare artificial intelligence (AI) evaluations, conceptualized as the specialized counterpart to ‘Gary the IT guy‘ but for medical and healthcare AI. Drawing inspiration from the Gary rule, which distinguishes between health IT and digital health based on whether a problem can be solved by ‘Gary, the IT guy‘ alone, MARK (Medical AI Reliability & Knowledge Framework) serves as your personal guide through the multifaceted process of healthcare AI assessment.
Created by healthcare and technology experts, MARK turns complex evaluation criteria into clear items, transparent scoring anchors and actionable recommendations. He doesn’t just identify problems. He shows you the evidence behind every score and helps you prioritize what to do next. Whether you are a clinician, healthcare administrator, researcher or innovator, MARK helps you judge the quality of AI-generated information before you rely on it.
What MARK 3.0 Evaluates (Information-Based Rubric, Research Mode)
| Domain | Items |
|---|---|
| Quality | Accuracy; Currency & Timeliness; Completeness; Logical Reasoning |
| Practicality | Relevance / Applicability; Usefulness |
| Reliability | Consistency; Transparency & Referral |
| Credibility | Attribution & References; Research Integrity & Safety; Balance & Objectivity |
| Clarity & Presentation | Language & Readability; Use of Numbers; Statistics & Data Presentation; Information Organization |
| User-Centeredness & Interaction | Personalization & Responsiveness to User Context; User Engagement; User Oversight & Responsibility |
Each item is scored from 0 to 3, for a maximum of 54 points. Scores are converted to percentages and banded as High Risk (0–33%), Moderate Risk (34–59%), Low Risk (60–79%) or Optimal (80–100%).
How MARK 3.0 Works
- Paste the AI output, or attach it (Word, PDF, text or a screenshot).
- Add the original question, the tool name and your intended audience.
- Optionally add a second output to test consistency.
- Run the evaluation. MARK scores 18 items across six domains.
- MARK builds your report. You get domain scores, strengths, weaknesses, prioritized recommendations for you and for the developer, and a list of what would raise confidence. Download it as a PDF, a Word file or JSON.
This content is only available to subscribers. Please register an account or login for access.
MARK Creators
- A/Prof. Kevin Yap
- Associate Professor, Alice Lee Centre for Nursing Studies, Yong Loo Lin School of Medicine, National University of Singapore
- Adjunct Associate Professor (Digital Health), School of Psychology and Public Health, La Trobe University
- A/Prof. Matthew Chua
- Adjunct Associate Professor, Department of Biomedical Informatics, Yong Loo Lin School of Medicine, National University of Singapore
- Mr. Adison Goh
- Data Scientist, National Healthcare Group, Digital Innovation Office, Ng Teng Fong Centre of Innovation, National Healthcare Group
MARK Team
The enhancement of MARK could not have been done without the help of the students who so enthusiastically took up their projects.
- Phuan Yun Xi & Zhang Zhen Hao (2025-2026)
- Project students from Department of Pharmacy and Pharmaceutical Sciences, National University of Singapore
- Ng Yi Sum & Mary Lim Shu Ying (2026-2027)
- Final Year Project students from Alice Lee Centre for Nursing Studies, Yong Loo Lin School of Medicine, National University of Singapore