摘要：The pioneering work of J.S. Haldane with the first decompression table in 1906 has generated considerable research and effort towards the development of safer and more rapid decompression procedures. The deterministic approach is governed by a fixed set of rules that defines the boundary between safe and unsafe dives and includes a model for gas exchange and an ascent criterion, such as gas supersaturation, to calculate the "safe" decompression depth. These decompression models are essentially empirical and provide "safe" decompression only over a limited range of depth and bottom times. The statistical approach considers DCI to be a probabilistic event and uses a risk function consisting of a gas exchange component and an ascent criterion to estimate or predict the risk of DCI. The ascent criterion can be based on supersaturation or bubble growth. To determine the risk function, a large data set of precise dive data, including time, depth, gas composition, and DCI outcome, must be available to match the predicted risk with the observed data. Probabilistic models of decompression can be used to analyze dive tables and procedures, compare different tables, and develop decompression tables with a given risk level. The probabilistic approach for decompression is a very powerful technique that could lead to a better insight into the physics and physiology of decompression because of its objectivity and potential for implementing a variety of models in the design of the risk functions for DCI. This review compares both approaches and discusses current and future challenges in the quest for a universal decompression model.
