Development of a Secure Web-Based Medical Imaging Analysis Platform: The AWESOMME Project.

Area of Science:

• Medical imaging analysis
• Machine learning in precision medicine
• Computational pathology

Background:

• Precision medicine relies on machine learning for processing patient data, including image analysis and computer-aided diagnosis.
• There is a significant need to effectively process and visualize large volumes of medical images and associated data.

Purpose of the Study:

• To propose an analysis pipeline for osteosarcoma patients using segmentation, feature extraction, and deep learning for treatment response prediction.
• To implement this pipeline on a secure, accessible web platform (AWESOMME project) for enhanced medical data analysis.

Main Methods:

• Development of a three-component web application architecture: data server, computation/authentication server, and a medical imaging web framework.
• Enhancement of existing components for security and traceability in expert data production.
• Integration of medical imaging processing steps: visualization, segmentation, feature extraction, and computer-aided diagnosis.

Main Results:

• The AWESOMME platform covers all medical imaging processing steps and facilitates the testing and use of machine learning models.
• The infrastructure is operational, deployed in internal production, and being installed in a hospital environment.
• User feedback and case study extensions have refined functionalities, demonstrating AWESOMME's modularity.

Conclusions:

• AWESOMME provides a modular solution for analyzing medical data and sharing research algorithms with clinicians.
• The platform enhances precision medicine research by enabling robust model creation and data processing for cancer treatment prediction.
• The implemented web application supports the continuous production of expert data through secure and traceable workflows.