Network blood flow simulation requires a
physiologically correct structural scheme of the individual vascular network.
Our goal is the development of (semi-)automatic MRI data processing
technology reducing the gap between raw data and computer simulations in clinic.

Two state-of-the-art software libraries provide such tools:
commercial code Amira and open source code VMTK. We use VMTK for 3D volume
extraction from individual MRI data for the vascular centerlines reconstruction.
For patient-specific vascular network reconstruction we adopt the open source
library VMTK to produce vascular centerlines on the basis of MRI data followed by
the developed automated skeletonisation algorithm. The produced vascular graph
possesses all necessary individual geometric data for 1D hemodynamic simulations.
The structural anatomical model and functional properties of the vessels
(elasticity and hydraulic resistance) were validated and fitted on the basis of
patient-specific MRI data and Doppler ultrasound measurements
provided by medical experts.

As a practical example of the personalized hemodynamic model we present
postsurgical haemodynamics analysis after femoral artery treatment of
occlusive vascular disease caused by atherosclerotic occlusion.
The analysis can be used as a personalized predictive
method  validated with clinical observations.

Other examples deal with non-invasive or minimally invasive procedures.
The non-invasive enhanced external counterpulsation procedure (EECP)
is used for stimulation of blood flow  in coronary vessels.
We developed a model of autoregulation impact on coronary flow during the EECP.
Minimally invasive procedures, such as cava-filter placement and artificial embolisation
of arterio-venous malformations, can be studied by our method of blood flow analysis as well.

computing methods; vascular network; cardiovascular modelling; 1D haemodynamics