🧬 AI-powered medical education

Learn medicine
through realistic synthetic clinical data,
without real patient risk.

MedEduSynth gives medical students hands-on experience with clinical data, AI diagnostics, and data-driven decision-making β€” using statistically realistic and clinically calibrated synthetic patients designed for safe educational simulation.

Request early access β†’ See how it works
6
Synthetic patients
18
Week curriculum
4
Learning phases
0
Coding required
MG
Maria G. β€” eGFR
38 ml/min Β· ↓ monitoring
Patient dashboard Β· Week 1
HbA1c 7.9% ↑
eGFR 38 ml/min ↓
Systolic BP 138 mmHg
Empagliflozin added βœ“ Month 9
15
Visits attended
3
Missed (MNAR)
18
Months data
Clinical decision
eGFR 32 β€” what do you recommend?
βœ“ Reduce Metformin + refer nephrology
6
TRL Level
Technology Readiness Level 6 β€” Advanced Prototype
MedEduSynth has been demonstrated in a relevant environment. The platform is not yet commercially available β€” we are currently seeking pilot partners for structured validation testing.
Apply for pilot access β†’
ANI β€” AgΓͺncia Nacional de Inovação
Financially supported by ANI
Aviso ANI-2025-01 Β· Programa Deep2start Β· FITEC
€60K
a fundo perdido
Voucher Deep Tech
Avanço na maturidade tecnológica (TRL).
Apoio a prototipagem, regulação e desenvolvimento de produto.
€10K
a fundo perdido
Voucher Go to EIC Accelerator
Apoio Γ  candidatura Full Proposal (Step 2)
do EIC Accelerator β€” Horizonte Europa.
ani.pt β†—
Regulatory & Data Protection Readiness

MedEduSynth is currently an advanced prototype at TRL 6, developed for medical education, clinical simulation and AI/data literacy training. In its current version, the platform is not intended to diagnose, treat, monitor or support clinical decisions on real patients.

All patient cases used in the platform are synthetic and designed for educational purposes. No real patient information is used, stored or processed in the generation or delivery of synthetic patient scenarios.

Pilot activities may involve the processing of user-related data, such as student progress, quiz results, submitted exercises, faculty feedback and anonymised performance metrics. These data are managed according to privacy-by-design and GDPR-readiness principles, with controlled access, role-based permissions and institutional oversight.

MedEduSynth is not currently CE-marked as a medical device and is not commercially available. The ongoing pilot programme is designed to validate educational performance, usability, faculty workflows, student engagement and responsible AI learning outcomes before future product-scale decisions.

How it works

Clinical thinking first.
Technology second.

Students never see a line of code. They reason clinically, explore data visually, and make simulated clinical decisions β€” the platform handles everything else.

01
πŸ“‹
Meet your synthetic patient
Each student receives a unique synthetic patient with a full clinical history β€” diagnoses, medications, and 18 months of realistic longitudinal data. No two students get the same case.
02
πŸ“Š
Explore & interpret data
Interactive charts, calendars, and simulators let students explore clinical trajectories without touching any code. They hover, toggle, and discover patterns β€” the way a clinician would.
03
🧠
Decide & communicate
Scenario-based simulated decisions at critical clinical moments, followed by writing real referral letters and clinical summaries. Assessed by faculty on clinical reasoning, not technical skills.
Synthetic patients

Six patients.
Six clinical stories.

All clinical stories are synthetic, educational and non-patient-specific. They are designed to reproduce realistic learning situations without reproducing or exposing real patient records.

Each patient is statistically calibrated against real clinical populations and designed to teach specific data reasoning skills β€” with hidden challenges built in.

MG
Maria G.
67F Β· High complexity
Type 2 Diabetes + CKD stage 3b. Teaches longitudinal EDA, missing data (MNAR), therapy change detection, and renal threshold decision-making.
Chronic Predictive Phase 1–4
CB
Carlo B.
52M Β· Acute urgency
STEMI + hypertension. Teaches high-frequency ICU data, ECG classification with neural networks, survival analysis, and algorithmic gender bias detection.
Acute ECG/Imaging Bias audit
AT
Amara T.
34F Β· Autoimmune
Systemic lupus erythematosus + anaemia. Teaches NLP on clinical reports, named entity recognition, and fine-tuning language models on Italian medical text.
Autoimmune NLP
LF
Luca F.
78M Β· Geriatric
Mild dementia + COPD + polypharmacy. Teaches multimorbidity data complexity, drug interaction modelling, and the limits of AI in elderly populations.
Geriatric Multimorbidity
SR
Sofia R.
28F Β· Obstetric
High-risk pregnancy with pre-eclampsia. Teaches time-series monitoring, early warning systems, and clinical alert design for critical obstetric events.
Obstetric Monitoring
YA
Yusuf A.
45M Β· Critical care
Post-surgical sepsis + septic shock. Teaches ICU streaming data, sepsis prediction models, and the ethical implications of automated early-warning systems.
Critical ICU data
Partnerships

Built on a foundation of European research collaboration.

MedEduSynth is developed by EUC Inovação Portugal β€” a Portuguese SME with an active portfolio of EU-funded projects in digital health, digital twins, and AI-based decision support. Our work is embedded in leading European research networks.

EU-funded R&D portfolio
Active participation in Horizon Europe Β· EIC Accelerator 2025 applicant (Proposal SEP-211231130). Recipient of two ANI financial vouchers: Voucher Deep Tech (€60K) and Voucher Go to EIC Accelerator (€10K) β€” Aviso ANI-2025-01, Programa Deep2start, funded by FITEC.
Health Cluster Portugal
Member of the Smart Health subcluster β€” a thematic network fostering collaboration in medical technologies and digital health.
EC Participant Register Β· PIC 917749205
Accredited by the Central Validation Service of the European Research Executive Agency for EU project participation.
We are looking for
πŸ›οΈ Medical schools & universities
Faculty interested in integrating data-driven clinical reasoning into their pre-clinical or clinical informatics curriculum.
πŸ₯ Hospitals & training centres
Clinical training departments seeking to upskill residents and junior doctors in AI-assisted decision support and health data literacy.
🀝 EU project consortia
Research consortia working on digital health, medical education, or AI ethics who need a validated synthetic patient education platform.
πŸ’Š Pharmaceutical & medtech companies
Industry partners interested in using synthetic patient platforms for training, research, or regulatory submission support.
Explore a partnership β†’
Letters of Intent Β· HORIZON-EIC-2025-ACCELERATOR-02 Β· Proposal ID SEP-211231130
Institutions that have expressed formal support

The following organisations have signed Letters of Intent supporting MedEduSynth within the EIC Accelerator 2025 application, recognising its potential to advance medical education and clinical AI training across Europe.

UniversitΓ  Politecnica delle Marche
Faculty of Medicine and Surgery Β· Italy
Signed by Prof. Mauro Silvestrini, Dean of the Faculty of Medicine. Supports MedEduSynth for medical education innovation using synthetic patients and digital twins.
univpm.it β†—
Maria Curie-SkΕ‚odowska University (UMCS)
Institute of Communication and Media Studies Β· Poland
Signed by Prof. Iwona Hofman (Director) and Prof. Ilona Biernacka-LigiΔ™za. Supports digital-twin and AI-enhanced education, student mobility and the Synthetic Patients Hub in Porto.
umcs.pl β†—
Health Cluster Portugal (HCP)
National Health Innovation Cluster Β· Portugal
Signed by Joaquim Cunha, Executive Director. HCP will support the dissemination of MedEduSynth among its members and partners to maximise European health ecosystem impact. Ref. 016/2025.
healthclusterportugal.pt β†—
CCG/ZGDV β€” Centro de Computação GrΓ‘fica
ICT Innovation Institute Β· Applied Research & Technological Innovation Β· Portugal
Signed by Ricardo J. Machado (Executive Chairman) and LuΓ­s Gonzaga MagalhΓ£es (Board of Trustees). Portuguese applied research institute in computer graphics and digital technologies β€” supporting simulation-based medical training and synthetic patient development.
ccg.pt β†—
These institutions have signed Letters of Intent in support of the MedEduSynth EIC Accelerator 2025 application Β· Proposal ID: SEP-211231130
18-week curriculum

Four phases.
One complete journey.

From reading a clinical file to deploying an AI diagnostic tool β€” all without requiring any prior programming experience.

1
Clinical Data Foundations
Weeks 1–4 Β· Patient: Maria G.
Students explore EHR structure, clean clinical datasets, perform longitudinal analysis, and visualise data trends β€” all through an interactive web interface, no coding required.
EDAMissing dataVisualisationStatistics
2
Predictive Modelling
Weeks 5–9 Β· Patients: Maria G. + Yusuf A.
Build and evaluate machine learning models for predicting clinical outcomes β€” CKD progression, sepsis risk, readmission. Interpret model performance using clinical metrics (AUC, sensitivity, specificity).
ML modelsROC curvesSHAPSurvival analysis
3
AI Diagnostics β€” NLP & Computer Vision
Weeks 10–14 Β· Patients: Carlo B. + Amara T.
Process clinical reports with NLP to extract diagnoses and therapy changes. Classify ECG signals and radiological images using neural networks. Explain AI decisions with Grad-CAM and LIME.
BioBERTECG classificationGrad-CAMmedSpaCy
4
Capstone Project & AI Ethics
Weeks 15–18 Β· Patient: assigned individually
Design and present an end-to-end clinical AI tool. Conduct a bias audit, write a Model Card, and present to a mixed jury of clinicians and data scientists. Includes AI Act and MDR regulatory framework.
End-to-end pipelineBias auditModel CardAI Act
Platform features

Built for medical students.
Not data scientists.

Every design decision prioritises clinical reasoning over technical complexity. Students think like doctors β€” the platform does the rest.

Statistically realistic synthetic patients
Each patient is generated using calibrated clinical distributions from real medical literature. Physiological correlations, realistic disease progression, and intentional data challenges are built in. Six exploratory patients available in this phase.
Zero coding required for students
Students interact through a clean web interface β€” interactive charts, clinical quizzes, decision scenarios, and referral letter writing. Python and AI run invisibly in the background, never exposed to learners.
Unique dataset per student
Every student receives a personalised patient dataset generated from their unique student ID. No two students analyse exactly the same data β€” eliminating copy-paste and fostering independent clinical reasoning.
Ethics & bias built into the curriculum
Algorithmic bias is not an afterthought β€” it is deliberately embedded in patient data. Students discover gender bias in triage models and must address it in their capstone project, aligned with EU AI Act requirements.
Faculty dashboard & analytics
Instructors see real-time student progress, submitted referral letters, quiz performance heatmaps, and the most common clinical reasoning errors across the cohort β€” without ever reviewing a line of code.
Privacy-by-design & pilot analytics
Pilot partners can access cohort-level educational analytics, student progress indicators and structured feedback tools. The platform is designed to support anonymised educational performance metrics, controlled access and institutional oversight.
GDPR-compliant by design
All data is synthetic β€” no real patient information is ever used, stored, or processed. The platform is designed for deployment in academic and hospital settings with full regulatory compliance from day one.
Pilot programme Β· TRL 6

Not for sale β€” yet.
But open for pioneers.

MedEduSynth is an advanced prototype (TRL 6), validated in a relevant environment. We are selecting a small cohort of pilot partners to test the platform in real academic settings and co-develop the final product.

Technology Readiness Level β€” MedEduSynth status
TRL 1
Basic principles
TRL 2
Concept
TRL 3
Proof of concept
TRL 4
Lab validation
TRL 5
Relevant env.
TRL 6 β—€
Demo prototype
WE ARE HERE
TRL 7
System demo
TRL 8
System complete
TRL 9
Market ready
What pilot partners receive
  • βœ“Full platform access for up to 50 students at no cost
  • βœ“All 6 synthetic patients + faculty dashboard
  • βœ“Direct access to the development team
  • βœ“Named acknowledgement in scientific publications
  • βœ“Priority access and preferential pricing at commercial launch
  • βœ“Co-design sessions to shape the final product roadmap
Partner commitments
  • β†’Run at least one full Phase 1 module with a student cohort (min. 10 students)
  • β†’Complete a structured feedback report within 30 days of pilot completion
  • β†’Allow collection of anonymised student performance metrics
  • * Pilot activities are limited to educational validation and do not involve clinical use on real patients.
  • β†’Participate in one 60-min video debrief session with the product team
  • β†’Provide a written reference or case study (LinkedIn / institutional testimonial)
No financial commitment. The pilot is free of charge. The only currency is structured, honest feedback that helps us build a better product together.
Apply for the pilot programme β†’

Currently accepting applications from medical schools, nursing schools, and health science faculties Β· Max 5 pilot partners in this cohort

"

The best way to learn clinical data reasoning is to get your hands on clinical data β€” safely, ethically, and at scale. MedEduSynth makes that possible for the first time.

VISION BEHIND MEDEDUSYNTH Β· Empowering Medical Learning Through Simulation
Get in touch

Ready to bring data-driven
medicine into your curriculum?

Whether you want to apply for the pilot programme, explore a partnership, or simply learn more β€” fill in the form and we'll get back to you within 2 working days.

What happens next?
1 We review your request within  2 working days
2 A 30-min intro call with our team
3 We tailor a pilot plan for your institution
Direct contacts
hello@mededusynth.eu +351 223 164 512
πŸŽ“ Max 5 pilot partners
Places allocated in order of application. Apply early.