From Research to Practice: Open AI Tools for Medicine
Medical Imaging Data for Research and AI Development
The AI4MED Image Archive provides a structured environment for accessing and exploring medical imaging data used in research, education, algorithm development, and validation. AI4MED contributes to the development of this research ecosystem by integrating medical imaging datasets with its activities in image processing, segmentation, quantitative analysis, simulation, and artificial intelligence.
The archive is intended to support researchers, students, medical physicists, imaging professionals, and AI developers in working with medical images within a common digital environment and in connecting imaging data with AI4MED-developed and integrated analytical workflows. This activity directly supports AI4MED projects dedicated to deep-learning medical image analysis and integration with the AI4MED database.
AI-Assisted Medical Image Analysis with MONAI-AL
MONAI-AL provides access to medical-image artificial intelligence workflows based on the MONAI ecosystem, supporting applications such as image segmentation, deep-learning experimentation, model development, and AI-assisted analysis.
AI4MED contributes by integrating MONAI-based technologies into its medical imaging research environment and applying them within its research activities, particularly in areas such as automated segmentation, tumor analysis, quantitative image assessment, and the evaluation of deep-learning methods. The integration provides researchers and students with an accessible environment in which modern AI methods can be explored in conjunction with medical imaging datasets and other AI4MED resources.
This implementation is aligned with AI4MED's broader research objectives of developing advanced methods for segmentation, classification, image reconstruction, predictive analytics, tumor detection, and quantitative imaging.
Radiation Dose and Radiobiological Risk Assessment
The AI4MED Dose & Radiobiological Risk Calculator is a research, educational, and decision-support application developed to estimate patient radiation exposure from diagnostic imaging examinations, including CT, conventional radiography, mammography, fluoroscopy, OPG, and CBCT.
The application can incorporate examination history, modality, anatomical region, repeated examinations, patient age and sex, and available technical dose parameters such as CTDIvol, DLP, KAP/DAP, kVp, mAs, and scan length to estimate effective and cumulative radiation dose. It also provides an evidence-based interpretation of potential stochastic radiobiological risk associated with cumulative ionizing-radiation exposure.
Unlike the other external frameworks presented in this section, this application represents a direct AI4MED research and development contribution. Integrated into the AI4MED platform, it is intended to support radiation-protection research, patient and professional education, protocol optimisation, and the future development of systematic patient-dose monitoring approaches
Online Visualization of Medical Images
The DICOM Viewer provides an accessible environment for viewing and examining medical images stored in the Digital Imaging and Communications in Medicine (DICOM) format. It supports the visualization and navigation of radiological image series and provides an important link between medical imaging data and subsequent processing or AI-assisted analysis.
Within the AI4MED ecosystem, the DICOM Viewer contributes to a broader workflow in which medical images can be accessed, visualized, examined and subsequently used for image-processing, segmentation, quantitative analysis, or AI-based research applications. AI4MED's contribution is therefore focused on integrating medical-image visualization into a unified research environment rather than developing the underlying DICOM standard itself.
This integration supports AI4MED's objective of creating practical pathways from medical imaging datasets to computational analysis, research, training, and translational applications
Advanced Medical Image Processing, Segmentation and 3D Analysis
3D Slicer is an established open-source platform for medical-image visualization, segmentation, processing, quantitative analysis, and three-dimensional reconstruction. Through the AI4MED environment, it is incorporated into research and educational workflows involving medical image analysis and AI-assisted applications.
AI4MED has used 3D Slicer in its medical imaging research, including work on MRI image segmentation and brain-image analysis. The AI4MED research programme extends this use toward workflows that combine advanced image-processing environments with deep-learning algorithms, quantitative feature extraction, tumor analysis, and medical-imaging simulation. AI4MED's official publication record specifically documents work on MRI image segmentation using the HD Brain Extraction extension in 3D Slicer.
Providing access to 3D Slicer through the AI4MED ecosystem supports researchers, students, medical physicists, and imaging professionals in moving from image visualization to segmentation, volumetric assessment, 3D reconstruction, and advanced quantitative analysis, while connecting these established tools with AI4MED research activities.
