Adaptive Inert Gas Exchange Model for Improved Hypobaric Decompression Sickness Risk Estimation
INTRODUCTION: Future high-altitude military operations and spaceflight will require new procedures to protect crews from decompression sickness while limiting the operational impact. It is hypothesized that the current prediction models do not accurately reflect actual inert gas dynamics, making them unsuitable for the risk estimation of new hypobaric exposure profiles. METHODS: A biophysical gas exchange model was created, allowing modification of various physiological parameters. Predicted nitrogen (N2) volume flows were compared with an experimental study by the Swedish Aerospace Physiology Centre. Bubble growth predictions, made using the Tissue Bubble Dynamics Model, were compared with measured venous gas emboli (VGE). RESULTS: While the simulated washout curves captured the general trends, some important discrepancies were observed when using the nominal model parameters. The new biophysical gas exchange model, incorporating changes in cardiac output and individual anthropometric variations, improved the predictions and approximated the experimentally observed N2 washout. The standard bubble growth predictions did not match measured VGE. Using weighing factors based on the N2 gas flow components predicted by the new biophysical model, the bubble growth pattern agrees much better with the measured VGE scores. DISCUSSION: Traditional decompression models do not account for variations in physiological and environmental factors, leading to incorrect estimates of N2 washout and bubble growth predictions. Using an adaptive biophysical gas exchange model significantly improves the predictions for various altitude exposure profiles. We therefore strongly recommend incorporating adaptive physiological parameters in any model to be used for estimating decompression sickness risk and designing mitigation procedures. De Ridder S, Neyt X, Germonpré P. Adaptive inert gas exchange model for improved hypobaric decompression sickness risk estimation. Aerosp Med Hum Perform. 2025; 96(2):85–92.
Basic structure of the adaptive biophysical gas exchange model. The human body is divided in six tissue compartments and a single alveolar compartment. Only two out of six body compartments are shown in the schematic. The heart is depicted by the RA (right atrium), RV (right ventricle), LA (left atrium), and LR (left ventricle). See text for symbol explanation (Methods section).
TOP: three different flight profiles used during simulated high-altitude flights in a hypobaric chamber. BOTTOM: measured 20-min average nitrogen volume flow values for the three conditions (gray boxes),8 compared to the model predictions using the nominal model parameters (solid line: continuous N2 flow prediction; hatched boxes: derived 20-min average values).
TOP: Decomposition of total N2 flow into flow components for Condition A, based on the compartments of the biophysical gas exchange model. BOTTOM: inert gas washout prediction for Condition A, using a 30% reduction in the cardiac output and the resulting derived first 20-min average value (hatched box).
TOP: decomposition of total N2 flow for the last intermittent recompression during Condition B. BOTTOM: total N2 flow during the last intermittent recompression of Condition B for two different people with different body characteristics (lean vs. heavyset).
LEFT: Median VGE scores in Conditions A, B, and C.8 CENTER: bubble growth prediction from the TBDM model. RIGHT: modified bubble growth prediction taking into account the dominating N2 flow compartments.
N2 flow pattern differences between Conditions B and C, focusing on the muscle and the fatty tissue compartment (compartments E and F, respectively).
Calculated bubble growth for each of the 10 compartments from the TBDM model, for Condition B and C. The 10 compartments have half-value times between 5–480 min. The solid line indicates the selected bubble growth curve, corresponding to the dominant N2 flow compartment (adipose tissue for Condition B and muscle tissue for Condition C).
Contributor Notes
