Personalised computational cardiology: Patient-specific modelling in cardiac mechanics and biomaterial injection therapies for myocardial infarction

Abstract

Predictive computational modelling in biomedical research offers the potential to integrate diverse data, uncover biological mechanisms that are not easily accessible through experimental methods and expose gaps in knowledge requiring further research. Recent developments in computing and diagnostic technologies have initiated the advancement of computational models in terms of complexity and specificity. Consequently, computational modelling can increasingly be utilised as enabling and complementing modality in the clinic-with medical decisions and interventions being personalised. Myocardial infarction and heart failure are amongst the leading causes of death globally despite optimal modern treatment. The development of novel MI therapies is challenging and may be greatly facilitated through predictive modelling. Here, we review the advances in patient-specific modelling of cardiac mechanics, distinguishing specificity in cardiac geometry, myofibre architecture and mechanical tissue properties. Thereafter, the focus narrows to the mechanics of the infarcted heart and treatment of myocardial infarction with particular attention on intramyocardial biomaterial delivery.

Keywords: Cardiac disease; Computational model; Finite-element method; Heart failure; Ischaemic heart disease; Subject specific.

Fig. 1

Number of yearly publications of peer-reviewed journal articles a with “patient-specific” or “subject-specific” contained in the title and b for finite-element-based studies focusing on cardiac ventricular mechanics. Source: Thomson Reuters ISI Web of Knowledge^® and PubMed^® databases, January 2016

Fig. 2

Patient-specific FE model for investigation of treatment of mitral valve regurgitation. Reproduced with permission from Baillargeon et al. [115]

Fig. 3

FE prediction of midwall fibre stress in an ovine left ventricle with anteroapical infarct without treatment (a) and with simulated intramyocardial delivery 4.4 mL of biomaterial in four infarct border zone locations indicated by arrows (b). Difference of midwall fibre stress between the untreated infarct and treated infarct that demonstrates the location of stress reduction in relation to the injection sites (arrows) (c). Adapted with permission from Wall et al. [95]

Fig. 4

Contour plots of fibre stress in the lateral wall of an ovine left ventricle with untreated infarct at end diastole (a) and end systole (b), and after treatment by delivery of 2.6 mL of a calcium hydroxyapatite-based tissue filler distributed over 20 evenly spaced injections at end diastole (c) and end systole (d). (Colour scales of the end diastole panels are the same, and colour scales of the end systole panels are the same). Adapted with permission from Wenk et al. [67]

Fig. 5

Ellipsoidal LV FE model with 20 intramyocardial hydrogel injectates. Adapted with permission from Kichula et al. [98]

Fig. 6

Histological micrographs demonstrating the distribution of a polyethylene glycol hydrogel (appearing in pink) delivered immediately (a) and seven days (b) after infarct induction in rat hearts (nuclei appear blue, bar represents 50 μm). Reproduced with permission from Kadner et al. [89]. Reconstructed 3D geometry of a polyethylene glycol hydrogel injectate with microstructural details reconstructed from histological sections in a biventricular rat heart geometry (c) (injectate shown in pink)

Fig. 7

Patient-specific LV FE model with 12 ellipsoidal hydrogel injectates located equidistant between the base and the apex. Adapted with permission from Lee et al. [104]