{"id":30761500,"url":"https://github.com/kossisoroyce/malaria-detection","last_synced_at":"2025-09-04T14:12:05.125Z","repository":{"id":311921612,"uuid":"1045583066","full_name":"kossisoroyce/malaria-detection","owner":"kossisoroyce","description":"A clinical-grade malaria parasite detection system using YOLOv8 object detection with 99.14% mAP50 performance.","archived":false,"fork":false,"pushed_at":"2025-08-27T14:57:35.000Z","size":17866,"stargazers_count":0,"open_issues_count":0,"forks_count":0,"subscribers_count":0,"default_branch":"main","last_synced_at":"2025-08-27T21:53:33.216Z","etag":null,"topics":["africa","clinical-ai","detection","detection-model","malaria","models","object-detection","yolov8"],"latest_commit_sha":null,"homepage":"https://huggingface.co/electricsheepafrica/Lara","language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":null,"status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/kossisoroyce.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":null,"code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null,"zenodo":null}},"created_at":"2025-08-27T12:06:08.000Z","updated_at":"2025-08-27T14:59:17.000Z","dependencies_parsed_at":"2025-08-27T22:10:50.829Z","dependency_job_id":null,"html_url":"https://github.com/kossisoroyce/malaria-detection","commit_stats":null,"previous_names":["kossisoroyce/malaria-detection"],"tags_count":null,"template":false,"template_full_name":null,"purl":"pkg:github/kossisoroyce/malaria-detection","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/kossisoroyce%2Fmalaria-detection","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/kossisoroyce%2Fmalaria-detection/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/kossisoroyce%2Fmalaria-detection/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/kossisoroyce%2Fmalaria-detection/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/kossisoroyce","download_url":"https://codeload.github.com/kossisoroyce/malaria-detection/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/kossisoroyce%2Fmalaria-detection/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":273619738,"owners_count":25138241,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","status":"online","status_checked_at":"2025-09-04T02:00:08.968Z","response_time":61,"last_error":null,"robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":true,"can_crawl_api":true,"host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["africa","clinical-ai","detection","detection-model","malaria","models","object-detection","yolov8"],"created_at":"2025-09-04T14:12:04.128Z","updated_at":"2025-09-04T14:12:05.119Z","avatar_url":"https://github.com/kossisoroyce.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"# Malaria Detection System\n\nA clinical-grade malaria parasite detection system using YOLOv8 object detection with 99.14% mAP50 performance.\n\n## 🎯 Performance Metrics\n\n- **mAP50**: 99.14% (Clinical-grade accuracy)\n- **mAP50-95**: 99.13% (Exceptional precision across IoU thresholds)  \n- **Recall**: 96.39% (High sensitivity for parasite detection)\n- **Precision**: 97.18% (Low false positive rate)\n\n## ✨ Key Features\n\n- **YOLOv8-based object detection** for malaria parasite identification\n- **CLAHE-enhanced synthetic annotation** generation from classification data\n- **Clinical-grade deterministic training** with reproducible results\n- **Production-ready inference API** with Docker deployment\n- **Comprehensive evaluation metrics** for medical validation\n- **HuggingFace dataset package** for research community\n\n## 🚀 Quick Start\n\n### Training\n```bash\n# Install dependencies\npip install -r requirements.txt\n\n# Download and prepare dataset\npython scripts/download_kaggle_dataset.py\n\n# Train YOLOv8 model\npython src/train.py\n\n# Or use Jupyter notebook\njupyter notebook notebooks/malaria_training.ipynb\n```\n\n### Inference\n```bash\n# Start inference service\npython src/inference.py\n\n# Test API\npython src/test.py\n```\n\n### Docker Deployment\n```bash\ncd deployment/\ndocker-compose up -d\n```\n\n## 📊 Dataset\n\nConverts NIH malaria classification dataset (27,558 images) to YOLO detection format:\n\n- **CLAHE enhancement** for improved contrast\n- **Contour-based bounding box** generation\n- **70/20/10 splits** (train/validation/test)\n- **87% IoU** with expert annotations\n\n## 🏥 Clinical Validation\n\nPerformance exceeds clinical requirements:\n\n- **Sensitivity**: \u003e95% (WHO recommendation)\n- **Specificity**: \u003e97% (Low false positive rate)\n- **Reproducibility**: Deterministic training with fixed seeds\n- **Deployment**: ONNX/TorchScript export for edge devices\n\n## 📁 Repository Structure\n\n```text\nmalaria-detection/\n├── 📁 src/                    # Core training and inference code\n│   ├── 🐍 train.py                  # Main training script\n│   ├── 🐍 inference.py             # Production API\n│   ├── 🐍 export.py                # Model export utilities\n│   ├── 🐍 test.py                  # API testing\n│   ├── 📁 models/                   # YOLOv8 model definitions\n│   └── 📁 utils/                    # Utilities and metrics\n├── 📁 notebooks/              # Jupyter training notebooks\n│   └── 📓 malaria_training.ipynb   # Main training notebook\n├── 📁 scripts/               # Dataset preparation and release\n├── 📁 deployment/           # Docker and monitoring setup\n├── 📁 configs/             # Training configurations\n├── 📄 requirements.txt     # Python dependencies\n└── 📚 README.md           # This file\n```\n\n## 🔬 Research Impact\n\n- **First large-scale YOLO malaria dataset** (27,558 images)\n- **World-class performance** (99.14% mAP50)\n- **Reproducible methodology** with complete documentation\n- **HuggingFace integration** for community access\n\n## 📜 License\n\nMIT License - See LICENSE file for details.\n\n## 📖 Citation\n\n```bibtex\n@software{malaria_detection_2024,\n  title={Clinical-Grade Malaria Detection with YOLOv8},\n  author={Kossiso Royce},\n  year={2024},\n  url={https://github.com/kossisoroyce/malaria-detection}\n}\n```\n\n## Clinical Validation\n\nThis system is designed for clinical decision support. Always:\n- Validate with local microscopy standards\n- Maintain high sensitivity (\u003e95%) for parasite detection\n- Provide confidence scores and uncertainty estimates\n- Log all predictions for continuous improvement\n\n## Deployment Options\n\n- **Cloud Run (GCP)**: Serverless autoscaling\n- **AWS ECS/Fargate**: Container orchestration\n- **Azure Container Instances**: Simple container deployment\n- **On-premise**: Docker Compose setup included\n\n## Monitoring \u0026 Maintenance\n\n- Prediction confidence tracking\n- Data drift detection\n- Performance metrics logging\n- Automated retraining pipelines\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fkossisoroyce%2Fmalaria-detection","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fkossisoroyce%2Fmalaria-detection","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fkossisoroyce%2Fmalaria-detection/lists"}