RT Journal Article SR Electronic T1 niiv: Fast Self-supervised Neural Implicit Isotropic Volume Reconstruction JF bioRxiv FD Cold Spring Harbor Laboratory SP 2024.09.07.611785 DO 10.1101/2024.09.07.611785 A1 Troidl, Jakob A1 Liang, Yiqing A1 Beyer, Johanna A1 Tavakoli, Mojtaba A1 Danzl, Johann A1 Hadwiger, Markus A1 Pfister, Hanspeter A1 Tompkin, James YR 2024 UL http://biorxiv.org/content/early/2024/09/13/2024.09.07.611785.abstract AB Three-dimensional (3D) microscopy data often is anisotropic with significantly lower resolution (up to 8×) along the z axis than along the xy axes. Computationally generating plausible isotropic resolution from anisotropic imaging data would benefit the visual analysis of large-scale volumes. This paper proposes niiv, a self-supervised method for isotropic reconstruction of 3D microscopy data that can quickly produce images at arbitrary (continuous) output resolutions. Within a neural field, the representation embeds a learned latent code that describes the implicit higher-resolution isotropic image region. Under isotropic volume assumptions, we self-supervise this representation on low-/high-resolution lateral image pairs to reconstruct an isotropic volume from low-resolution axial images. We evaluate our method on simulated and real anisotropic electron (EM) and light microscopy (LM) data. Compared to a state-of-the-art diffusion-based method, niiv shows improved reconstruction quality (+1 dB PSNR) and is over three orders of magnitude faster (2,000×) to infer. Specifically, niiv reconstructs a 1283 voxel volume in 1/10th of a second, renderable at varying (continuous) high resolutions for display.Competing Interest StatementThe authors have declared no competing interest.