NanoMorph-3D: An End-to-End Physics-Driven Unrolling Framework for Nanomaterial Reconstruction

2026-08-04

AI Summary

This work addresses the geometric distortions in three-dimensional reconstructions of nanomaterials caused by missing wedge artifacts and noise in electron tomography. The authors propose an end-to-end physics-driven unrolling network that integrates a learnable proximal gradient descent architecture with a hierarchical attention mechanism to capture long-range 3D dependencies. A dual-domain sinusoidal attention module is specifically designed to explicitly model projection trajectories. Furthermore, an unsupervised dual-stream alignment strategy combined with physics-based normalization effectively bridges the domain gap between simulated and real data. Extensive experiments across diverse nanomaterial topologies demonstrate that the proposed method significantly enhances reconstruction fidelity, suppresses missing wedge artifacts, and achieves faster computational performance.

Abstract

Precise 3D characterization of nanomaterials is essential for unlocking structure-property relationships. However, standard electron tomography is fundamentally limited by the missing wedge problem. Consequently, conventional algorithms suffer from severe geometric distortions, a challenge further complicated by pervasive noise interference. Current learning-based methods either rely on physics-blind post-processing or employ end-to-end architectures constrained by local receptive fields, failing to capture complex 3D topologies. We propose NanoMorph-3D, a unified end-to-end framework grounded in a comprehensive Nanomorphological Taxonomy. Powered by a large-scale synthetic dataset explicitly modeling non-linear electron attenuation, we design a Physics-Driven Unrolled Network mapping proximal gradient descent into a learnable architecture. To capture complex internal topologies, we formulate a hierarchical attention mechanism with Physics-Normalization for long-range 3D dependencies and scale invariance. Crucially, our Dual-Domain strategy leverages Sinusoidal Attention to explicitly model physical projection trajectories, enforcing strict sinogram consistency to mitigate missing wedge artifacts. Finally, an unsupervised dual-stream mechanism bridges the simulation-to-reality gap. Experiments demonstrate NanoMorph-3D reconstructs diverse topologies with superior fidelity and speed.

Problem

• nanomaterial reconstruction
• missing wedge problem
• 3D topology
• electron tomography
• geometric distortion

Innovation

• Physics-Driven Unrolling
• Hierarchical Attention
• Sinusoidal Attention
• Missing Wedge Mitigation
• Dual-Domain Reconstruction