Subject Specific Left Ventricular Modelling

Grand challenges lie in the subject-specific left ventricular modelling, as shown in Fig.1

Subject specific modelling challenging with in-vivo data

          Fig. 1 Challengings related to subject specific LV modelling from in-vivo clinical data

To tackle those challengings, we are developing towards personalized left ventricular modellings as proposed in Fig.2

Glasgow Heart Components

                                  Fig.2 main components of subject-specific left ventricular modelling

Several major steps for constructing a subject specific model 

  • Step 1: Clinical in-viva data, shown in Fig.3

 in vivo cmr images

                                                   Fig.3 in-vivo CMR images

  • Step 2: Image processing, involving segmentation, rigid registration, non-rigid registration, quantification ane so on

LV model Reconstruction

                                     Fig.4 illustration of computational LV geometry reconstruction

     deformable image registration

                Fig.5 deformable image registration for strain recovery from cine MRI

oedema scar segmentation

                                      Fig.6 semi-automatic oedema/scar region quantification

IBFE brief introduction

                                               Fig.7 Brief introduction of immersed boundary method

       Several ways to incorporate layered myofibers into LV model with mapping ex-vivo data base shown in Fig.8 or rule-based method. Currently in-vivo data is extremely hard to acquire.

           myofibre architecture

                                      Fig.8 Myofiber architecture reconstruction from ex-vivo DT-MRI

  • Material models (passive and active)

H-O passive model

                                Fig.9 orthotropic passive myocaridal material model (structure-based)

Active material model

                               Fig. 10 active contraction model in the cellular level

  • Material parameter identification (inverse problem)

inverse estimation of material parameters

                                                    Fig.11 multi-step optimization procedure

  • Step 4: Validation

     Fig.12 shows the simulated LV endocardial surface is overlaped with a long axis CMR image from a healthy volunteer from end of diastole to end of systole

Deformed LV endo surface superimposed on a long axis image

                                           Fig.12 simulated LV endocardial surface (red)

Fig.13 is the strain comparison between simulated results (red) and mri measurements (blue)  from end of diastole to end of systole

strain comparison for a healthy volunteer

                                                                                  Fig.13 strain comparison

Videos and Presentations

LV model with inflow and outflow tracts


The corresponding LV endocaridal surface dynamics superimposed on the MR images

Presentation in the 7th World Congress of Biomechanics: