medical-imaging-ai

What problem does it solve?

This Skill provides a comprehensive framework for designing, building, validating, deploying, and monitoring AI/ML models on medical images, addressing the complexities of medical imaging AI development.

Core Features & Use Cases

• Medical Imaging AI Development: Supports all stages of AI/ML model development for medical imaging.
• DICOM Processing: Offers guidance on DICOM preparation, anonymization, and preprocessing.
• Data Sources: Includes information on various data sources, including public datasets and institutional PACS/VNA.
• Model Architecture: Recommends model architectures based on the task and data size.
• Evaluation: Provides guidelines for model evaluation, including metrics and subgroup analysis.
• Deployment: Offers insights into deployment patterns, including orchestrators and architecture choices.
• Regulatory Compliance: Assists with regulatory framing, including SaMD risk classification and FDA submissions.
• Post-Deployment Monitoring: Provides guidelines for monitoring and maintaining deployed models.

Quick Start

Use the medical-imaging-ai skill to design an AI model for chest CT nodule detection and integration into the radiologist's workflow.