A new biophysical decompression model for estimating the risk of articular bends during and after decompression

The biophysical models that intend to predict the risk of decompression sickness after a change of pressure are not numerous. Few approaches focus in particular on joints as target tissues, with the aim to describe properly the mechanisms inducing pain. Nevertheless, for this type of decompression incidents, called articular bends, no model proved to fit the empirical results for a broad range of exposures and decompression procedures. We present here an original biophysical decompression model for describing the occurrence of articular bends. A target joint is broken down into two parts that exchange inert gases with the blood by perfusion and with each other by diffusion over distances of a few millimetres. This diffusion pathway allows the slow amplification of microbubbles growing during and after decompression, consistent with the possible delayed occurrence of bends. The diffusion coefficients introduced into this model are larger than those introduced into most modern decompression models. Their value remains physical (#10(-9)m(2)/s). Inert gas exchanges and the formation, amplification and resorption of microbubbles during and after decompression were simulated. We used a critical gas volume criterion for predicting the occurrence of bends. A risk database extracted from COMEX experience and other published studies were used for the correlation of model parameters not known a priori. We considered a large range of exposure, and the commonly used inert gases nitrogen and helium. This correlation phase identified the worst biophysical conformations most likely to lead to the formation, in tissues such as tendons, of a large number of microbubbles recruited from pre-existing gas nuclei during decompression. The risk of bends occurrence was found to be linked to the total separated gas volume generated during and after decompression. A clamping phenomenon occurs soon after the start of decompression, greatly slowing the gas exchanges controlled especially by the oxygen window. This model, which reproduces many empirical findings, may be considered both descriptive and predictive.