An evaluation framework for sparse 4D (3D + time) imaging reconstruction via bootstrapped cross‐validation
Open Access
- 1 November 2026
- journal article
- Published by Wiley in Journal of Microscopy
Abstract
Four‐dimensional (4D; 3D + time) microscopic imaging has emerged as a powerful technique for investigating dynamic phenomena in complex systems, enabling direct visualisation of structural evolution in space and time. However, when pushing the limits of spatiotemporal resolution, most time‐resolved imaging techniques yield inherently sparse 4D datasets. While deep learning‐based reconstruction methods have shown promise in reconstructing 4D from sparse spatiotemporal measurements, a practical approach for evaluating their performance in the absence of a 4D reference has, to the best of our knowledge, been lacking. Here, we present a bootstrapped cross‐validation framework that estimates reconstruction performance by quantifying correlations between reconstructions generated from independently sampled subsets of the acquired data, as inspired by the 3D validation strategy in cryo‐electron microscopy, where reconstructions from split datasets are compared to assess resolutions. This enables both qualitative and quantitative assessment in the absence of ground truth. We investigate two representative scenarios with sparse and ultra‐sparse X‐ray datasets and validate this approach using 4D‐ONIX, a 4D deep‐learning reconstruction method, on simulated water droplet collision experiments. The proposed approach provides a reference‐free framework for performance estimation and support for better‐informed experimental strategies across a wide range of ultrafast imaging applications.Keywords
Funding Information
- European Research Council (ERC‐2020‐STG3DX‐FLASH948426)
This publication has 64 references indexed in Scilit:
- Super Time‐Resolved TomographyAdvanced Science, 2025
- NerT-CA: Efficient Dynamic Reconstruction from Sparse-View X-ray Coronary Angiography2025
- CrossScore: Towards Multi-View Image Evaluation and Scoring2024
- High-Speed, 3D Volumetric Displacement and Strain Mapping in Soft Materials Using Light Field MicroscopyExperimental Mechanics, 2022
- Tomoscopy: Time‐Resolved Tomography for Dynamic Processes in MaterialsAdvanced Materials, 2021
- Stereoscopic PIV measurements using low-cost action camerasExperiments in Fluids, 2021
- High resolution volumetric dual-camera light-field PIVExperiments in Fluids, 2019
- Particle Image Velocimetry2018
- I See What You Do Not See2017
- Real time volumetric ultrasound imaging systemJournal of Digital Imaging, 1990