Analyzing gene expression patterns over the tissue of an organ during growth is a means to decipher the biological mechanisms that control its shaping. For this, the organ is often sampled and imaged in different individuals at different developmental stages. However, estimating a continuous developmental process from a discrete dataset is an arduous task, especially if both size and shape evolve. We propose a strategy to: 1) integrate individual data to produce a comprehensive representation of the dynamic of expression patterns in a growing tissue, and 2) superimpose and compare patterns coming from different experimental groups. This opens the way to the integrative atlasing of expression patterns in an organ during its development. To illustrate our approach, we use a particularly challenging organ that can encompass huge shape changes during growth, the plant leaf.