This special issue, edited by Irene E. Vignon-Clementel et al, was published in the Annals of Biomedical Engineering.
The special issue is a collection of selected works presented at the Virtual Physiological Human 2020 (VPH2020) conference.
The VPH conference series, which focuses on Computational Systems Biomedicine, aims to make in silico medicine a reality. To this end, an essential step is the development of mathematical models that describe relevant physiological and/or pathological processes. For example, in this issue Guerrero et al proposed a model of cardio-respiratory interactions for obstructive sleep apnea. In line with the increased attention in recent years on the microvasculature, Spieker et al. investigated fundamental mechanisms of platelet adhesion at the single vessel scale. Modeling vascularized tissues presents specific challenges as vessel networks essentially represent connected segments within a 3D block of tissue. Multiscale model strategies are derived to account for the effect of blood flow pressure on the surrounding elastic tissue or to investigate oxygen transport.
In order for in silico medicine to become a reality, the methods developed to solve the model equations need to be efficient. For example, Pfaller et al proposed a methodology to monitor and achieve faster blood flow periodicity of the solution using a coarser scale initial solution. Identifying what model refinement is needed depending on the desired output is also important for the clinical application, as demonstrated for the case of brain perfusion by Josza et al A key component of the application to the clinical setting is model parameterization from patient data and the associated uncertainty quantification. Gillette et al developed an automated personalization strategy of their cardiac electrophysiology model based on combined MRI-ECG data. Antonuccio et al studied the uncertainty of model output based on patient-specific data for coarctation of the aorta.
Experiments and imaging are key constituents of in silico model parameterization and/or validation. Machine learning algorithms are increasingly used for experimental image analysis, as was done here by Sheriff et al for blood platelets. Annio et al validated a CFD-enhanced imaging method through in vitro experiments. Another approach is the in silico testing of experimentally-observed phenomena, as demonstrated by Lindsey et al to better understand the mechanobiology relationship between hemodynamics and congenital heart defect morphogenesis.
VPH2020 had a strong focus on clinical applications: better understanding or monitoring diseases, planning treatment, and assessing risks. For example, the in silico biomechanics model of Viceconti et al provides insight into the impact of suboptimal neuromuscular control on the risk of knee implant failure. Combining in silico biomechanics modelling and in vivo imaging, Narang et al can predict before surgery, the ischemic mitral valve regurgitation recurrence. The conference also highlighted newer clinical application areas such as reproduction and pregnancy. For instance, Fidalgo et al investigated scarring due to a Cesarean section through a computational biomechanics model of delivery. Finally, in silico trials are emerging as a powerful tool as shown by Favre et al who provided a perspective on in silico clinical trials for orthopedic devices, highlighting challenges and potential benefits.