Quantifying neuronal differentiation using temporal topological persistence

Neuronal anatomical differentiation relies on coordinated neurite dynamics to establish class-type morphology during development. Recent advances in high-throughput time-lapse imaging techniques have transformed our ability to track such growth dynamics, yielding comprehensive anatomical datasets of neuronal morphologies at unprecedented rates. However, analyzing these complex datasets using traditional morphometrics requires manual feature selection, which leads to costly and biased quantifications. To overcome these limitations, we applied the Topological Morphology Descriptor (TMD) to extract topological representations of neurons from time-lapse imaging of two cell types within the well-established Drosophila larval sensory dendritic arborization (da) system. We first used TMD to accurately classify the developmental trajectory of Class I da sensory neurons across embryonic and larval stages. However, TMD alone failed to distinguish class-type-specific morphology in Class III neurons, particularly in detecting altered branching rate dynamics in genetically modified mutants. To address this, we extended TMD to a Temporal Topological Morphology Descriptor (TTMD) to incorporate temporal dynamics. TTMD successfully classified mutations in Drosophila Class III da neurons, capturing altered branching rate dynamics in mutant phenotypes. These findings highlight the power of TTMD as an unbiased, scalable framework for analyzing neuronal growth dynamics and linking structural development to genetic and functional variation.