MSD knowledge management system for CAD

MultiScaleHuman (MSH) project goal is to create multi-scale biomedical data simulation, visualization, and knowledge management systems for improved understanding, diagnosis and treatment of musculoskeletal diseases (MSD) and related disorders. Within the MultiScaleHuman (MSH) project scope, the MultiScaleHuman Knowledge Management (MSH-KM) system targets to offer various tools and services that can be built on the top of MSH ontology (see deliverable D8.2 PM 36, D8.3 PM 48) for efficient management, sharing and retrieval of biomedical data and information. The knowledge management system aims to facilitate an integrated analysis of physiological human articulation that can offer computer guided assistance in the diagnosis and treatment of MSD. The MultiScaleHuman knowledge management platform (MSH-KM) is not a single, but a combination of correlated tools/techniques, which are guided by the MSH ontology. In the previous edition of the deliverable MSH D8.4 (PM36), we depicted preliminary developments related to the standalone annotation client and the web-hosted platform for biomedical data sharing and browsing. In this version, we revise the former designs and introduce new solutions to support more specific applications focused on the diagnosis and treatment of musculoskeletal diseases. Our objective is to go beyond building a mere data storage platform, and develop a computational framework that takes into account the morphological and structural information of patient-specific data, and integrates them within the formalized bio-medical domain knowledge. This aim can be considered as a possible contribution to a modern clinical decision support system. Whereas the data sets available to the MultiScaleHuman Consortium are acquired only from the healthy volunteers, a pathological case-study is required to fulfil our research objectives and validate the real potential of MSH-KM modules, especially the computational components. Thanks to the collaboration with other projects [1] [2], we formulated a clinically validated case-study for the diagnosis and follow-up study of Rheumatoid Arthritis (RA), a chronic inflammatory musculoskeletal disorder, focusing on carpal district. Carpal district is one of the most complex joints in the human body, since it is a collection of multiple bones and joints. All of these bones participate in complex articulations that allow variable mobility of the wrist. Creating a computer-guided assistance in the early stage diagnosis and treatment of RA on carpal district is considered as an interesting research challenge in the medical domain. To support our case-study, the MSH ontology has been extended with the formalization of carpus (c.f. Sec. 4.1 and D 8.3 PM 48). We used a training dataset of carpus provided by MSH Partner Softeco Sismat that contains 30 healthy and 18 pathological 3D bone models manually segmented from T1 weighted MRI images of wrist district. Additionally, we run some preliminary validation checks with "The digital database of wrist bone anatomy" where the 3D models are segmented from CT images [3]. We also executed some preliminary analysis on the healthy bone models of the knee region acquired and segmented within the MSH project scope. In the previous version of this deliverable D8.4 "MSD knowledge management system for CAD" (PM 36), we specified a detailed description of this requirement analysis, including some user scenarios, and offered the core services of the MSH-KM system based on the derived Competency Questions. In the current deliverable, we have revised the older version of MultiScaleHuman Knowledge Management (MSH-KM) [version D8.4 (PM36)] and are extending the scope further to support our research objectives. The ultimate target is to validate the MSH-KM suitability in real clinical data management scenario, and combine the MSH-KM services with the functionality of a standalone CAD system in order to propose significant improvements in the traditional medical investigation pipeline. As envisaged in D8.4 (PM 36), in this version of the deliverable D8.4 (PM 48) we are depicting the MSH-KM with the core components implemented: ?MSH-KM standalone annotation client: SemAnatomy3D - sets up a concrete bridge between the formalisation of the clinically relevant information related to the medical data and their geometric representations. ?MSH knowledge-base: MSH ontology and SemAnatomy3D annotation data model - manage the storage of annotated 3D models according to the annotation data model and the MSH ontology schema for efficient retrieval. ?MSH-KM: data sharing and semantic search-services for sharing and retrieval of the data and information present in the MSH knowledge base, leveraging biomedical knowledge formalization in the MSH ontology.