Deep Learning for Brain Age Estimation: Bridging Neuroimaging and Human Factors Engineering

Recent advancements in deep learning have significantly improved brain age classification from MRI, offering valuable insights for Human Factor Engineering (HFE). This review highlights state-of-the-art models, includ-ing multi-task autoencoders, adversarial variational networks, and mul-timodal 3D CNNs, which achieve high accuracy in estimating chronologi-cal and biological brain age. These models often integrate structural and functional MRI, along with synthetic data like AI-generated cerebral blood volume, to enhance performance. Techniques such as Grad-CAM and SHAP improve interpretability, revealing critical brain regions asso-ciated with aging. Reported mean absolute errors under 3 years demon-strate strong predictive capabilities. These methods support age-specific system design by assessing cognitive and physiological readiness. By link-ing neuroimaging and HFE, brain age prediction enables more adaptive, user - centred engineering solutions. This review synthesizes key methods, results, and trends to guide future research in this interdisciplinary do-main.