KMSS: A Knowledge Management System for Senology

In this paper, we propose the design and development of a Knowledge Management System for Senology (breast cancer domain) called KMSS. It will include all information and knowledge derived from the patient records and imaging modalities such as mammography, echography, MRI (Magnetic Resonance Imaging), scanner and ultrasound. Knowledge management (KM) is the process of capturing, developing, sharing, and effectively using organizational knowledge. Before starting the development of the KMSS system, we have analyzed the users' needs using a requirement analysis methodology. We use the Unified Modeling Language (UML) to specify and express these requirements. It allows us to summarize all concepts and senologic knowledge. These concepts are formalized with various 'scenarios' and 'use cases'. This formalization in natural language may help senologist neophytes in computer science to understand both the analysis and design of the system. We have developed an ontology of the domain by standardizing senological concepts and constraints using the BI-RADS (Breast Imaging Reporting and Data System) Atlas. We have represented these concepts as cases according to the case-based reasoning (CBR) approach. The KMSS system will be implemented in the department of Radiology of Gustave Roussy Institute in Paris (France). It will be interfaced with other existing radiological systems such as HIS (Hospital Information System), RIS (Radiological Information System) and PACS (Picture Archiving and Communication System) and will be compliant with commercial systems for senology such as CAD (Computer-Aided Diagnosis), BI-RADS and DICOM (Digital Imaging and COmmunication in Medicine). Details concerning the design and development of the KMSS system will be described in the paper.