The most common solid tumour found in children is Neuroblastoma (NB), but the mechanisms that lead to its formation are still not fully understood. This study aims to gain a better understanding of how spheroid growth evolves over time and how it is regulated by mechanical stimuli. The study proposes a complete workflow that begins by describing experiments in which NB tumour individual cells were cultured in 3D microfluidic devices upon collagen-type I hydrogels. The spheroids' evolution was monitored for seven days, during which the cross-sectional area of the spheroids demonstrated non-linear growth [1]. To simulate spheroid growth, a multiphase poroelastic model based on porous media was employed to accurately account for the features of the experimental set-up. The model used a poroelastic approach, which is composed of a solid part (extracellular matrix) and two fluid phases (tumour cells and interstitial fluid) to represent the collagen network of the hydrogel. The multiphase model had already been formulated and validated in the literature. To better understand how tumour growth was evolving, the study aimed to estimate the main mechanical parameters of the model. A sensitivity analysis based on Sobol indices was used to identify the most influential model parameters. Five of the twenty-one parameters were found to be the most relevant with respect to the spheroid volume. An inverse analysis based on Bayesian techniques was then used to infer these functions towards the posterior density function, based on prior knowledge of the influential parameters from literature values. The resulting posterior distribution demonstrated that the multiphase computational model was able to reproduce the spheroid growth observed in vitro and to replicate the experimental variability. As a result, the study was able to estimate tumour mechanical properties, such as the spheroid permeability and the dynamic viscosity of tumour cells, which are not feasible to measure in the experimental set-up [4].