Self-Supervised Super-Resolution Approach for Isotropic Reconstruction of 3D Electron Microscopy Images from Anisotropic Acquisition

TL;DR

This paper introduces a self-supervised deep learning method using a vision-transformer architecture to reconstruct isotropic 3D electron microscopy images from anisotropic data, improving analysis and visualization without needing paired training datasets.

Contribution

A novel self-supervised deep learning framework with ViT blocks for isotropic 3DEM reconstruction from anisotropic data, eliminating the need for paired datasets.

Findings

Successfully reconstructed isotropic 3DEM from anisotropic data

Outperformed existing methods in quality of reconstruction

Validated on three brain datasets

Abstract

Three-dimensional electron microscopy (3DEM) is an essential technique to investigate volumetric tissue ultra-structure. Due to technical limitations and high imaging costs, samples are often imaged anisotropically, where resolution in the axial direction ($z$) is lower than in the lateral directions $(x,y)$. This anisotropy 3DEM can hamper subsequent analysis and visualization tasks. To overcome this limitation, we propose a novel deep-learning (DL)-based self-supervised super-resolution approach that computationally reconstructs isotropic 3DEM from the anisotropic acquisition. The proposed DL-based framework is built upon the U-shape architecture incorporating vision-transformer (ViT) blocks, enabling high-capability learning of local and global multi-scale image dependencies. To train the tailored network, we employ a self-supervised approach. Specifically, we generate pairs of anisotropic and isotropic training datasets from the given anisotropic 3DEM data. By feeding the given anisotropic 3DEM dataset in the trained network through our proposed framework, the isotropic 3DEM is obtained. Importantly, this isotropic reconstruction approach relies solely on the given anisotropic 3DEM dataset and does not require pairs of co-registered anisotropic and isotropic 3DEM training datasets. To evaluate the effectiveness of the proposed method, we conducted experiments using three 3DEM datasets acquired from brain. The experimental results demonstrated that our proposed framework could successfully reconstruct isotropic 3DEM from the anisotropic acquisition.