Coupled experimental and computational approach to stomach biomechanics: Towards a validated characterization of...

Gastric diseases are one of the most relevant healthcare problems worldwide. Interventions and therapies definition/design mainly derive from biomedical and clinical expertise. Computational biomechanics, with particular regard to the finite element method, provides hard-to-measure quantities during in-vivo tests, such as strain and stress distribution, leading to a more comprehensive and promising approach to improve the effectiveness of many different clinical activities. However, reliable finite element models of biological organs require appropriate constitutive formulations of building tissues, whose parameters identification needs an experimental campaign consisting in different tests on human tissues and organs. The aim of the reported here research activities was the identification of mechanical properties of human gastric tissues. Human gastric specimens were tested at tissue, sub-structural and structural levels, by tensile, membrane indentation and inflation tests, respectively. On the other hand, animal experimentations on tissue layers from literature pointed out the mechanical response at sub-tissue level during tensile loading conditions. In detail, the analysis of experimental results at sub-tissue and tissue levels led to a fibre-reinforced visco-hyperelastic constitutive formulation and to the identification of gastric layers mechanical behaviour. Results from experimentations on human samples were coupled with data derived from animal models. Data from sub-structural and structural experimentations were exploited to upgrade and validate the constitutive formulations and parameters. The developed investigations led to a reliable constitutive framework of human gastric tissues that both describe stomach mechanical functionality and allow computational investigations. Indeed, the comparisons among average computational data and experimental medians provided the following RMSEs (Root Mean Square Errors): 0.89 N, 0.15 N for corpus and fundus during membrane indentation test, respectively, and 0.44 kPa during inflation test. Accounting for the magnitude of experimental and computational data, the RMSEs can be considered low and acceptable because they concerned biological samples. In fact, biological tissues and structures are affected by a high inherent inter-samples' variability, which is detectable in both the geometrical configuration and the mechanical behaviour. The specific values of the here reported RMSEs ensured the reliability of the achieved parameters and the quality of the overall developed procedure. Reliable computational models of the gastric district could become efficient clinical tools to find out the main crucial aspects of bariatric procedures, such as the mechanical stimulation of gastric mechano-receptors. Moreover, the methods of computational biomechanics will permit to run the preliminary tests of new and innovative bariatric procedures, on one hand, predicting the successful rate and the effectiveness, and, on other hand, reducing the use of animal testing.