Meda Knowledge Core bridges raw PubMed biomedical literature, GRADE evidence-based knowledge graphs, and an AI-augmented course authoring workspace with pluggable local & cloud LLMs.
# Prise en charge initiale de l'AVC Ischémique aigu
L'administration d'altéplase IV (rtPA) dans la fenêtre des 4h30 réduit significativement le handicap fonctionnel à 3 mois (mRS 0-1).
"Générer 3 questions de validation clinique de niveau Bloom 4 sur les contre-indications..."
Ingests PubMed literature and breaks down complex clinical findings into n-ary Statements connecting Class A–E Atoms, bounded by clinical and epidemiological contexts.
Advanced dual-pane Markdown & PDF course creator with real-time token highlighting, dynamic Table of Contents, and multi-format DOCX/PDF exports.
Directly plug into your model of choice — whether cloud LLMs via OpenRouter (Claude 3.5, GPT-4o) or 100% private local models running on Ollama & Llama.cpp.
Select or highlight any clinical term to query systematic reviews, meta-analyses, forest plot statistics (RR, OR, I²), and insert verified citations on demand.
MKKC-V3 implements an 8-tier role architecture natively supporting Country Admins, Deaneries, Faculty Professors, and Residents.
Configure regional clinical guidelines (HAS, WHO, NICE), ICD-11 ontology baselines, and pedagogical criteria across universities.
Oversee medical faculties, hospital departments, residency quotas, and curriculum progress with unified RBAC.
Author clinical courses with the Phodoc markdown editor, multi-LLM brainstorming (OpenRouter & local Ollama), and PubMed meta-analysis linking.
Audit automatically extracted Atoms and Statements, verifying GRADE evidence levels and clinical confidence scores.
Empower medical professors to brainstorm, draft curricula, and synthesize complex clinical guidelines with complete freedom over their AI infrastructure:
Indexed in 4 Systematic Reviews with 9 Meta-Analyses. Absolute risk reduction of 9.8% for disability at 90 days.