{"ID":2874313,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2509.05238","arxiv_id":"2509.05238","title":"Uncertain but Useful: Leveraging CNN Variability into Data Augmentation","abstract":"Deep learning (DL) is rapidly advancing neuroimaging by achieving state-of-the-art performance with reduced computation times. Yet the numerical stability of DL models -- particularly during training -- remains underexplored. While inference with DL is relatively stable, training introduces additional variability primarily through iterative stochastic optimization. We investigate this training-time variability using FastSurfer, a CNN-based whole-brain segmentation pipeline. Controlled perturbations are introduced via floating point perturbations and random seeds. We find that: (i) FastSurfer exhibits higher variability compared to that of a traditional neuroimaging pipeline, suggesting that DL inherits and is particularly susceptible to sources of instability present in its predecessors; (ii) ensembles generated with perturbations achieve performance similar to an unperturbed baseline; and (iii) variability effectively produces ensembles of numerical model families that can be repurposed for downstream applications. As a proof of concept, we demonstrate that numerical ensembles can be used as a data augmentation strategy for brain age regression. These findings position training-time variability not only as a reproducibility concern but also as a resource that can be harnessed to improve robustness and enable new applications in neuroimaging.","short_abstract":"Deep learning (DL) is rapidly advancing neuroimaging by achieving state-of-the-art performance with reduced computation times. Yet the numerical stability of DL models -- particularly during training -- remains underexplored. While inference with DL is relatively stable, training introduces additional variability prima...","url_abs":"https://arxiv.org/abs/2509.05238","url_pdf":"https://arxiv.org/pdf/2509.05238v1","authors":"[\"Inés Gonzalez-Pepe\",\"Vinuyan Sivakolunthu\",\"Yohan Chatelain\",\"Tristan Glatard\"]","published":"2025-09-05T16:54:26Z","proceeding":"math.NA","tasks":"[\"math.NA\",\"cs.AI\"]","methods":"[\"Convolutional Neural Network\"]","has_code":false}