Regional Registration of Whole Slide Image Stacks Containing Major Histological Artifacts

Abstract

Background: High resolution 2D whole slide imaging provides rich information about the tissue structure. This information can be a lot richer if these 2D images can be stacked into a 3D tissue volume. A 3D analysis, however, requires accurate reconstruction of the tissue volume from the 2D image stack. This task is not trivial due to the distortions such as tissue tearing, folding and missing at each slide. Performing registration for the whole tissue slices may be adversely affected by distorted tissue regions. Consequently, regional registration is found to be more effective. In this paper, we propose a new approach to an accurate and robust registration of regions of interest for whole slide images. We introduce the idea of multi-scale attention for registration. Results: Using mean similarity index as the metric, the proposed algorithm (mean +- std: 0.84 +- 0.11) followed by a fine registration algorithm (0.86 +- 0.08) outperformed the state-of-the-art linear whole tissue registration algorithm (0.74 +- 0.19) and the regional version of this algorithm (0.81 +- 0.15). The proposed algorithm also outperforms the state-of-the-art nonlinear registration algorithm (original: 0.82 +- 0.12, regional: 0.77 +- 0.22) for whole slide images and a recently proposed patch-based registration algorithm (patch size 256: 0.79 +- 0.16 , patch size 512: 0.77 +- 0.16) for medical images. Conclusion: Using multi-scale attention mechanism leads to a more robust and accurate solution to the problem of regional registration of whole slide images corrupted in some parts by major histological artifacts in the imaged tissue.

Extraction and Visualization of Ocular Blood Vessels in 3D Medical Images Based on Geometric Transformation Algorithm

“…Since SIFT feature descriptors determine the location of key points, the SIFT algorithm determines the feature points in an image based on the grayscale distribution in the image. e extracted feature descriptors contain both geometric and grayscale information [20]. e feature points it extracts are unique and the feature points are robust to image translation, scaling, rotation, radiometric distortion, a large amount of noise, and illumination changes, making it possible to extract feature points for correct matching in a large amount of image data.…”

Reconhecimento de artefatos durante a análise de lâminas histológicas: uma revisão integrativa

“…Além disso, a normalização sequencial das imagens gera artefatos estruturais (Zheng, 2019;Izadyyazdanabadi, 2019;Yagi, 2016). Uma análise 3D, no entanto, requer uma reconstrução precisa do volume do tecido da pilha de imagens 2D, porém segundo Paknezhad, deve-se considerar a forte presença de deformações de tecido nas fatias virtuais adquiridas e o grande tamanho dessas imagens, assim, os métodos baseados em pontos de referência podem ser mal orientados por regiões altamente deformadas no tecido (Paknezhad, 2020).…”

Three‐dimensional tissue volume generation in conventional brightfield microscopy

“…This technology has been used for many applications during the last 20 years ranging from studying the anatomy of embryos (Sharpe, 2003) to visualizing the morphology of cells (Belay et al, 2021). Another alternative for 3D tissue volume reconstruction has been presented for whole slide imaging stacks using registration algorithms (Paknezhad et al, 2020). However, the technologies that have been presented in literature (Belay et al, 2021; Davis et al, 2019; Du et al, 2020; Liu et al, 2019; Magsam et al, 2018; Nguyen et al, 2017; Prunskaite‐Hyyrylainen, 2019; Sharpe, 2003; Vallejo Ramirez et al, 2019; Zhang et al, 2020) a/ require the sample to be fully rotated around the light source, b/ necessitate the design of a special instrument devoted for this purpose and c/ have limited application in clinical practice.…”