Predictions from a mathematical model of decompression compared to Doppler scores.

This paper describes an attempt to calibrate a mathematical model that predicts the extent of bubble formation in both the tissue and blood of subjects experiencing decompression from a hyperbaric exposure. The model combines an inert gas dynamics model for uptake and elimination of inert anesthetic gases with a simple model of bubble dynamics in perfused tissues. The calibration has been carried out using the model prediction for volume of free gas (bubbles) as microl/ml in central venous blood and relating this to Doppler scores recorded at the end of hyperbaric exposures. More than 1,000 Doppler scores have been compared with the model predictions. Discriminant analysis has been used to determine the cut-points between scores below a certain level and all scores at or above that level. This allows each prediction from the model to be equated to a particular pattern of bubble scores. The predictions from the model are thus given a context against the more familiar Doppler scores as a means of evaluating decompression stress. It is thus possible to use the mathematical model to evaluate decompression stress of a hyperbaric exposure in terms of the predicted volume of gas that will form into bubbles and to convert that to a prediction of the most likely pattern of Doppler grades which would be recorded from a group of subjects experiencing that exposure. This model has been used in assisting regulators to set limits to the level decompression risk that should be considered acceptable and in assisting those working with decompression procedures to design effective modifications.