Medical Imaging & DICOM Analysis

Analyze medical imaging data from DICOM servers and Picture Archiving and Communication Systems (PACS). This skill enables AI agents to query imaging studies by patient, modality, or date range; retrieve and parse DICOM metadata for clinical context; process diagnostic imaging reports; and support radiological workflows including study comparison, modality-specific analysis, and structured reporting -- bridging the gap between imaging data systems and clinical decision support.

Quick Install

npx skills add Open-Medica/open-medical-skills --skill medical-imaging-analysis

What It Does

• PACS/DICOM server querying using C-FIND and DICOMweb (QIDO-RS) to search for imaging studies by patient ID, accession number, modality type (CT, MRI, US, XR, NM, PET), body part, study date range, or referring physician
• Study-level metadata analysis extracting study descriptions, series counts, image counts, acquisition parameters (slice thickness, field strength, contrast administration), and radiation dose information (CTDIvol, DLP) when available
• Multi-modality comparison support enabling side-by-side analysis of studies across different time points or modalities (e.g., comparing a baseline CT with a follow-up PET/CT to assess treatment response)
• Structured radiology report generation from DICOM metadata and study parameters, producing standardized report templates with findings organized by anatomical region, impression summaries, and follow-up recommendations using ACR-compliant formatting
• Imaging protocol compliance verification to check whether acquired studies match the ordered imaging protocol specifications (contrast phase, slice thickness, coverage extent, reconstruction kernel)

Clinical Use Cases