SP3 - Pathway & statistical modeling

Predictive & Bio-Inspired Modeling Group, Dr. Sven Hirsch Group (ZHAW) 

This subproject will formulate the model that links the mechanically related biological findings (SP2) with the fluid and structure mechanical model (SP4). This subproject will integrate known and new information across observation scales and time to generate a dynamic model of the disease. In this manner we can predict the altered composition of the aneurysm wall needed to simulate the structural changes. The local modification in wall function will inevitably lead to change in 3-dimensional shape. This fundamental tool will be validated against observation in growing aneurysms in animal side-wall experiments (SP2) and selected clinical cases (SP1) 

Secondly, this group will be responsible to develop the statistical disease model together with the clinicians of SP1. For this we will first understand how the aneurysm can be best represented with shape descriptors using the established 3D Zernike invariant moments as a starting point. We then develop metrics to compare different aneurysm shapes. The last step is to set up a machine-learning based Bayesian framework to explore causal chains within the database. This exploratory tool will be handed to the clinicians to further improve the disease model.

Function:

1. Parameter fitting of pathway model (with SP2).

2. In silico experiments to verify and validate FSG model with animal studies (with SP2&5).

3. Establish similarity metrics based on 3D shape descriptors.

4. Exploratory tool for multivariate disease prediction & classification (with SP1&5). 

Expected added value to systems medicine: 

A refined statistical and pathway model of the disease.