Logo Lanfrica
  • Accueil
  • Atlas
  • Analyses
  • Documentation
  • Sign in

© 2026 Lanfrica. Tous droits réservés. Tous les droits d'auteur des ressources affichées sur le site Web Lanfrica appartiennent aux détenteurs de droits d'auteur d'origine, sauf indication contraire explicite.

mrselemogo777-gif/african-fracture-ai

Domaine:

healthcare

Type de record:

modeldataset
Créateur:
mrs
Hôte:
ML-powered fracture triage system for rural African clinics. Uses clinical vitals + X-ray analysis with 84.1% accuracy. Trained on 1,129 patients from 18 African countries. # African Fracture AI ML-powered fracture triage system for rural African clinics. Uses clinical vitals + X-ray analysis with 84.1% accuracy. Trained on 1,129 patients from 18 African countries. --- ## 📊 Model Performance ### Clinical Model (Random Forest) | Metric | Value | |--------|-------| | Accuracy | 81.4% | | Cross-validation (5-fold) | 82.4% (±1.7%) | | Precision (Simple) | 81.4% | | Recall (Simple) | 79% | | Precision (Complex) | 83% | | Recall (Complex) | 89% | ### X-ray Model (EfficientNetB0 + PCA + RF) | Metric | Value | |--------|-------| | Accuracy | 67.7% | | Precision | 71.3% | | Recall | 54.3% | | Confusion Matrix | Simple: 98 correct, 23 false alarm; Complex: 57 correct, 48 missed | ### Multimodal Ensemble (Late Fusion) | Metric | Value | |--------|-------| | Accuracy | 82.4% | | Clinical Weight | 91% | | X-ray Weight | 9% | | Test Set Size | 46 aligned patients | | True Negatives | 22 | | False Positives | 3 | | False Negatives | 4 | | True Positives | 17 | ### Feature Importance (Clinical Model) | Feature | Importance | |---------|------------| | Pain Scale | **58.7%** | | Heart Rate | 10.8% | | Age | 10.2% | | Systolic BP | 10.2% | | Diastolic BP | 10.1% | **Interpretation:** Pain scale dominates prediction, which aligns with clinical expectation that patient-reported pain is the strongest indicator of fracture severity. --- ## 📸 Screenshots ### Web Application Interface **Homepage** **Model Input Form** **Results and Prediction** **Feature Importance Analysis** **Country Data Visualization** **Model Performance Metrics** --- ## 🌍 18 African Countries Represented | Country | Patients | Country | Patients | |---------|----------|---------|----------| | Tanzania | 96 | Kenya | 65 | | South Africa | 76 | Uganda | 63 | | Ghana | 75 | Mali | 61 | | Zambia | 75 | Egypt | 59 | | Burkina Faso | 72 | Ethiopia | 59 | | Zimbabwe | 68 | Ivory Coast | 57 | | Senegal | 66 | Nigeria | 52 | | | | Botswana | 51 | | | | Cameroon | 47 …

Visit

github.com

Similaires

mrselemogo777-gif/ML-Assisted-Fracture-Triage-for-Rural-African-Clinics-A-Comparative-Study-mrselemogo777-gif/tlhagiso-backendcreova-gif/kilimo-aigrandelagbanou28-gif/AgriBen-AI

mrselemogo777-gif/ML-Assisted-Fracture-Triage-for-Rural-African-Clinics-A-Comparative-Study-

# ML-Assisted Fracture Triage for Rural African Clinics: A Comparative Study ## Abstract This study

mrselemogo777-gif/tlhagiso-backend

Tlhagiso Backend API - SMS spam detection and URL phishing detection for Botswana and the World

creova-gif/kilimo-ai

KILIMO AI - AgTech for East Africa # Kilimo AI **An AI-assisted farming companion for East African

grandelagbanou28-gif/AgriBen-AI

Premium Flutter app for African farmers — crop management, plant disease detection, marketplace & ag