Polymeric foams are widely used in various industries due to their lightweight nature, high strength-to-weight ratio, and excellent energy absorption properties [1]. Understanding and predicting their complex microstructural behaviour under mechanical loading is crucial for optimising their performance in engineering applications. This study presents a novel approach combining in-situ X-ray computed tomography with convolutional neural networks to model the mechanical response of polymeric foam materials. In-situ X-ray computed tomography enables non-destructive, three-dimensional imaging of the foam’s evolving microstructure during deformation, capturing critical features such as cell wall buckling, collapse, and fracture. The high-resolution volumetric data serve as rich input for training convolutional neural network architectures designed to learn spatial patterns and correlations linked to mechanical behaviour. By leveraging convolutional neural networks’ capability to extract hierarchical features automatically, the proposed method provides an efficient and accurate predictive model without relying on traditional, computationally expensive finite element simulations. The developed convolutional neural network framework successfully predicts stress-strain outputs and failure for any foam microstructure, validated against experimental deformation data. This data-driven approach facilitates rapid assessment of material performance under diverse loading conditions, enabling accelerated design and optimisation of polymeric foams with tailored mechanical properties. This integration of advanced imaging and deep learning techniques represents a significant advancement in multiscale material modelling by bridging experimental observations with predictive analytics. The findings demonstrate the potential of in-situ X-ray computed tomography based convolutional neural networks as a powerful tool for material scientists and engineers aiming to innovate lightweight, high-performance polymeric foams.