Understanding how biological systems spontaneously generate spatial patterns is fundamental to developmental biology. While quantitative frameworks for positional information have been successful in simple model organisms like Drosophila, extending these approaches to mammalian systems has been limited by technical challenges in imaging dense tissues and extracting single-cell information. This thesis establishes a comprehensive methodological framework for quantifying gene expression patterns in gastruloids, self-organizing stem cell aggregates that recapitulate key aspects of mammalian embryonic development. An optimized optical clearing protocol was developed that enables complete 3D confocal immunofluorescence imaging of these dense structures while preserving morphology and fluorescence. Combined with state-of-the-art deep learning segmentation, this pipeline successfully extracts single-cell information throughout entire gastruloid volumes.