TY - JOUR
T1 - Cognitive and physiological assessment in a Mars Analog Habitat
T2 - Implications for aerospace medical monitoring in hypobaric hypoxia
AU - Huang, Yuran
AU - Ding, Li
AU - Pei, Jiye
AU - Li, Yan
N1 - Publisher Copyright:
© 2025
PY - 2026/2
Y1 - 2026/2
N2 - With the advancement of deep space exploration, the analog cabin pressure environment proposed for future Mars missions (8.0 psia, 32 % O2) has attracted increasing attention, as it enables the convenient execution of various simulation experiments. In this study, we employed wearable multimodal physiological monitoring and mental workload (MWL) assessment, designed to operate under low-pressure and mildly hypoxic environments. Utilizing the OpenMATB task paradigm, we recorded electroencephalography (EEG), electrocardiography (ECG), electromyography (EMG), and peripheral oxygen saturation (SpO2) to examine how hypoxia and task difficulty jointly influence MWL dynamics. Ten healthy participants performed seven-level OpenMATB tasks under three simulated altitude conditions—ground level (GL: 40 m), moderate altitude (MA: 2400 m), and high altitude (HA: 3000 m)—while also completing NASA-TLX assessments. Results revealed that as altitude increased, SpO2 significantly decreased, accompanied by mild autonomic adjustments and relative motor-area β power in EEG. Nevertheless, Mental Demand scores of NASA-TLX were significantly higher under hypoxic conditions (p = 0.047), indicating greater perceived cognitive effort in reduced-pressure environments, while behavioral performance remained stable, suggesting short-term compensatory adaptation. In contrast, task level exerted significant effects on all NASA-TLX dimensions (p < 0.001), partial task performance, and EEG and HRV indices, indicating that workload rather than hypoxia was the primary factor driving physiological modulation. Machine learning models that leveraged 2-min time windows, Lasso-based feature selection, and Random Forest (RF) or eXtreme Gradient Boosting (XGBoost) classifiers achieved up to 96.3 % accuracy in classifying the seven MATB workload levels and yielded R2 values approaching 0.91 in predicting continuous NASA-TLX scores, confirming the consistency between classification and regression performance. Overall, short-term moderate hypoxia mainly affected peripheral oxygenation and perceived workload, whereas cortical and autonomic functions were predominantly shaped by task difficulty. These findings provide a technical foundation for adaptive workload monitoring in future partially pressurized habitats and long-duration space missions.
AB - With the advancement of deep space exploration, the analog cabin pressure environment proposed for future Mars missions (8.0 psia, 32 % O2) has attracted increasing attention, as it enables the convenient execution of various simulation experiments. In this study, we employed wearable multimodal physiological monitoring and mental workload (MWL) assessment, designed to operate under low-pressure and mildly hypoxic environments. Utilizing the OpenMATB task paradigm, we recorded electroencephalography (EEG), electrocardiography (ECG), electromyography (EMG), and peripheral oxygen saturation (SpO2) to examine how hypoxia and task difficulty jointly influence MWL dynamics. Ten healthy participants performed seven-level OpenMATB tasks under three simulated altitude conditions—ground level (GL: 40 m), moderate altitude (MA: 2400 m), and high altitude (HA: 3000 m)—while also completing NASA-TLX assessments. Results revealed that as altitude increased, SpO2 significantly decreased, accompanied by mild autonomic adjustments and relative motor-area β power in EEG. Nevertheless, Mental Demand scores of NASA-TLX were significantly higher under hypoxic conditions (p = 0.047), indicating greater perceived cognitive effort in reduced-pressure environments, while behavioral performance remained stable, suggesting short-term compensatory adaptation. In contrast, task level exerted significant effects on all NASA-TLX dimensions (p < 0.001), partial task performance, and EEG and HRV indices, indicating that workload rather than hypoxia was the primary factor driving physiological modulation. Machine learning models that leveraged 2-min time windows, Lasso-based feature selection, and Random Forest (RF) or eXtreme Gradient Boosting (XGBoost) classifiers achieved up to 96.3 % accuracy in classifying the seven MATB workload levels and yielded R2 values approaching 0.91 in predicting continuous NASA-TLX scores, confirming the consistency between classification and regression performance. Overall, short-term moderate hypoxia mainly affected peripheral oxygenation and perceived workload, whereas cortical and autonomic functions were predominantly shaped by task difficulty. These findings provide a technical foundation for adaptive workload monitoring in future partially pressurized habitats and long-duration space missions.
KW - Deep space mission
KW - Hypobaric hypoxia
KW - Machine learning
KW - Mars habitat
KW - Mental workload
KW - Multimodal physiological signals
KW - OpenMATB
KW - Wearable monitoring system
UR - https://www.scopus.com/pages/publications/105021033187
U2 - 10.1016/j.actaastro.2025.11.002
DO - 10.1016/j.actaastro.2025.11.002
M3 - 文章
AN - SCOPUS:105021033187
SN - 0094-5765
VL - 239
SP - 294
EP - 306
JO - Acta Astronautica
JF - Acta Astronautica
ER -