Models and simulation techniques for discovering diabetes influence factors

Challenges

Technological Objectives

The main technological challenge of MOSAIC is to develop and improve existing empirical models to perform a better diagnose, prognosis and prediction of complications in people at risk of suffering from T2DM or already diagnosed. The following technological objectives represent the solutions of the strategic approaches identified in the Scientific Objectives.

TO1: Improve existing modelling and prediction tools in diabetes management

MOSAIC will develop new models adopting combined bottom-up and top-down approaches:

1- Microscopic clinical insight will be propagated upwards through the organs level, integrated at individual level and at population level.
2- Information gathered through time-series and process based models will be propagated downward by extracting knowledge from high level datasets and resulting in simpler models.

a) Discovering novel biomarkers through extended probabilistic models à understanding cellular mechanisms beyond glucose regulation through more comprehensive and new multi-scale modelsà enhance type 2 Diabetes onset and evolution through empowered predictive models.

b) Extending TAR methodology to study the temporal evolution of diabetes population. A research direction that will be pursued in MOSAIC is the joint analysis of clinical and behavioural information (potentially derived from administrative data, such as drug consumptions, and GP visits). To this end, we will work on a class of methods the MOSAIC partners have already applied on Diabetes data.

c) Process Mining.

TO2: Accelerated diagnosis and Characterization of type 2 diabetes through health technologies devices and tools

Resulting models in MOSAIC will also assess the value and performance of monitoring technologies, like continuous glucose monitoring technologies, physical activity monitoring devices, patient’s digital assistant used to collect monitoring, therapeutic and environmental data, as innovative tools to identify people at risk of type 2 diabetes, enabling intervention to prevent or delay the onset of type 2 diabetes with the final objective of designing a comprehensive model for integrated care of type 2 diabetic patients through synchronized strategies of continuous Metabolic Assessment Continuous Stratification Program, Risk detection and Intervention program based on reference works.

TO3: Integrate models in current decision support systems and enhance decision-making in clinical care

The MOSAIC models will result in computerized tools that will be tested in a clinical environment. To achieve this goal, these tools will be integrated in a diabetes management system that will act as the vehicle to facilitate the usage of the tools in a clinical environment. The consortium has at his disposal two diabetes disease management systems, METABO (FP7 – 216270) already validated in a clinical trial and NOHMAD Chronic, which is a certified Chronic Disease Management platform commercialized by TSB, that guarantees the highest level of security, accuracy, interoperability with the legacy systems in the clinical trials sites, and accountability.

The tools developed within MOSAIC have the potential to be used in the next generation of clinical decision support applications enhancing the current systems, which are usually based in static rules rather than in intelligent data mining and modeling and, therefore, do not have the capability to adapt to the specific conditions and risks of patients following real life challenges and changes.

MOSAIC will follow a subdivision of tasks into different WPs to delineate accurately the scope of the work to be carried out in each of them.

The three main technical WPs, WP2, WP3 and WP4, will be responsible for the technical developments of the MOSAIC prototypes. WP1 will support all these activities aligning them with the scientific and clinical objectives of the project and enabling the use of high quality pools of data for the multi parametric analysis.

WP6 and WP7 cover the horizontal activities and support the technical WPs along the project. WP5 will integrate all the activities related to the validation of the concepts, algorithms, models and prototypes produced along the life of the project.

It is fundamental keep a close relationship and interaction among all WPs. Even if they keep the essence of the development of a particular component and/or activity in the system, they are interrelated as it is shown in the following figures.