Zebrafish Histotomography Noise Removal In Projection And Reconstruction Domains

Adishesha, Amogh Subbakrishna; Vanselow, Daniel J.; Rivière, Patrick La; Huang, Xiaolei; Cheng, Keith C.

doi:10.1109/isbi48211.2021.9433914

Cited by 1 publication

(1 citation statement)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Two immediate approaches for decreasing the scan times are available: a) Reducing the X-ray exposure time at each of these angles and denoising the resulting scans as demonstrated in [38], [41] and [6] or b) Reducing the number of sampling angles around the subject as performed in [11] and [27]. In the former technique, since fewer X-ray photons reach the sensor, there is an increase of Poisson-based noise in the scans that can severely deteriorate the quality of the scans and can obscure biologically relevant sub-cellular details.…”

Section: Introductionmentioning

confidence: 99%

Sinogram Domain Angular Upsampling of Sparse-View Micro-CT with Dense Residual Hierarchical Transformer and Noise-Aware Loss

Adishesha

Vanselow

Rivière

et al. 2023

Preprint

Self Cite

View full text Add to dashboard Cite

Reduced angular sampling is a key strategy for increasing scanning efficiency of micron-scale computed tomography (micro-CT). Despite boosting throughput, this strategy introduces noise and artifacts due to undersampling. In this work, we present a solution to this issue, by proposing a novel Dense Residual Hierarchical Transformer (DRHT) network to recover high-quality sinograms from 2x, 4x and 8x undersampled scans. DRHT is trained to utilize limited information available from sparsely angular sampled scans and once trained, it can be applied to recover higher-resolution sinograms from shorter scan sessions. Our proposed DRHT model aggregates the benefits of a hierarchical- multi-scale structure along with the combination of local and global feature extraction through dense residual convolutional blocks and non-overlapping window transformer blocks respectively. We also propose a novel noise-aware loss function named KL-L1 to improve sinogram restoration to full resolution. KL-L1, a weighted combination of pixel-level and distribution-level cost functions, leverages inconsistencies in noise distribution and uses learnable spatial weights to improve the training of the DRHT model. We present ablation studies and evaluations of our method against other state-of-the-art (SOTA) models over multiple datasets. Our proposed DRHT network achieves an average increase in peak signal to noise ratio (PSNR) of 17.73dB and a structural similarity index (SSIM) of 0.161, for 8x upsampling, across the three unique datasets, compared to their respective Bicubic interpolated versions. This novel approach can be utilized to decrease radiation exposure to patients and reduce imaging time for large-scale CT imaging projects.

show abstract