# Numerical Uncertainty of Convolutional Neural Networks Inference for Structural Brain MRI Analysis

arXiv: https://arxiv.org/abs/2308.01939 · HTML: https://mlsanity.com/paper/2308.01939/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 3 Aug 2023.

## Built on by (most cited first, 2 of 2)

- [Fuzzy PyTorch: Rapid Numerical Variability Evaluation for Deep Learning Models](https://mlsanity.com/paper/2605.25991/) (2026)
- [Uncertain but Useful: Leveraging CNN Training Variability into Data Augmentation](https://mlsanity.com/paper/2509.05238/) (2025)

## Abstract (arXiv)

This paper investigates the numerical uncertainty of Convolutional Neural Networks (CNNs) inference for structural brain MRI analysis. It applies Random Rounding -- a stochastic arithmetic technique -- to CNN models employed in non-linear registration (SynthMorph) and whole-brain segmentation (FastSurfer), and compares the resulting numerical uncertainty to the one measured in a reference image-processing pipeline (FreeSurfer recon-all). Results obtained on 32 representative subjects show that CNN predictions are substantially more accurate numerically than traditional image-processing results (non-linear registration: 19 vs 13 significant bits on average; whole-brain segmentation: 0.99 vs 0.92 Sorensen-Dice score on average), which suggests a better reproducibility of CNN results across execution environments.