Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2024 Jan 17;14(1):1497.
doi: 10.1038/s41598-024-52007-5.

Full resolution reconstruction of whole-mount sections from digitized individual tissue fragments

Daan Schouten  1 , Jeroen van der Laak  2 , Bram van Ginneken  3 , Geert Litjens  2
Affiliations

Full resolution reconstruction of whole-mount sections from digitized individual tissue fragments

Daan Schouten et al. Sci Rep. .

Abstract

Whole-mount sectioning is a technique in histopathology where a full slice of tissue, such as a transversal cross-section of a prostate specimen, is prepared on a large microscope slide without further sectioning into smaller fragments. Although this technique can offer improved correlation with pre-operative imaging and is paramount for multimodal research, it is not commonly employed due to its technical difficulty, associated cost and cumbersome integration in (digital) pathology workflows. In this work, we present a computational tool named PythoStitcher which reconstructs artificial whole-mount sections from digitized tissue fragments, thereby bringing the benefits of whole-mount sections to pathology labs currently unable to employ this technique. Our proposed algorithm consists of a multi-step approach where it (i) automatically determines how fragments need to be reassembled, (ii) iteratively optimizes the stitch using a genetic algorithm and (iii) efficiently reconstructs the final artificial whole-mount section on full resolution (0.25 µm/pixel). PythoStitcher was validated on a total of 198 cases spanning five datasets with a varying number of tissue fragments originating from different organs from multiple centers. PythoStitcher successfully reconstructed the whole-mount section in 86-100% of cases for a given dataset with a residual registration mismatch of 0.65-2.76 mm on automatically selected landmarks. It is expected that our algorithm can aid pathology labs unable to employ whole-mount sectioning through faster clinical case evaluation and improved radiology-pathology correlation workflows.

PubMed Disclaimer

Conflict of interest statement

J.L. is chief scientific officer and shareholder of Aiosyn. G.L. is shareholder of Aiosyn and member of the advisory board of Canon health informatics. However, none of these positions were related to or had any impact on the presented work.

Figures

Figure 1
Figure 1
Comparison of (a) whole-mount sectioning and (b) quartered sectioning for a transversal cross-section of a prostatectomy specimen.
Figure 2
Figure 2
Example of the solution ranking from two prostatectomy cases with four fragments with the correct solution annotated in green. All solutions were scored and ranked on the mean registration error between stitch edges before any optimization with the genetic algorithm. Overlapping areas between adjacent fragments are indicated in white. The top row shows an example where the correct solution had the lowest registration error and thus the highest rank. The bottom row shows an example where the correct solution was ranked fourth due to a substantial registration error originating from significant overlap between the upper left and lower left quadrant.
Figure 3
Figure 3
Fragment configuration accuracy when only taking the top N ranked solutions into account, shown for both the case (a) without manual correction and (b) with manual correction. This figure aims to demonstrate the trade-off between increased accuracy and the approximate linear increase in computational overhead with every additionally included solution.
Figure 4
Figure 4
Violin plot of the absolute registration error for internal test set 4 as measured on a rough initial initialization and four different resolution levels. The labels on the x-axis indicate by which factor the original WSI was downsampled where "initial" refers to the rough first stitch initialization without optimization by the genetic algorithm at a downsampling factor of 2560.
Figure 5
Figure 5
Representative example case for each test set showing the individual fragments and the final stitched result.
Figure 6
Figure 6
Schematic overview of the entire reconstruction pipeline for an example prostatectomy case with four fragments. Compartment (a) represents the automatic fragment configuration where the stitch edges of each fragment are extracted, all possible configurations (64 in the case of 4 fragments) are computed and evaluated to find the top N best solutions. Compartment (b) displays how the top N solutions are finetuned with a genetic algorithm and emphasizes the multi-resolution iterative aspect. Compartment (c) demonstrates the tile-based approach where the final full-resolution reconstruction is computed tile by tile.
Figure 7
Figure 7
Schematic overview of the proposed blending method. A and B show a small tile of two individual fragments which pose a slight overlap in the final reconstruction. C displays the region of overlap and the gradient that is computed to facilitate the gradient blending as described in Eq. (3). D shows a magnified version of the area marked by the red box in A-C where the final image is computed by averaging the pixel intensities from fragment A and B. The red arrows indicate how this results in clear boundary artefacts at the edge of the overlapping area. E shows our proposed gradient blending method where the boundary artefacts seem visually mitigated.
Figure 8
Figure 8
Example of a whole-mount (left) which was artificially cut into four different fragments. Note how the edges along which the fragments were cut are not fully straight and mimic the slightly curved edges of the fragments obtained from the quartered sectioning as displayed in Fig. 1.
See this image and copyright information in PMC

Similar articles

See all similar articles

References

    1. Foschini MP, Baldovini C, Ishikawa Y, Eusebi V. The value of large sections in surgical pathology. Int. J. Breast Cancer. 2012;1–7:2012. doi: 10.1155/2012/785947. - DOI - PMC - PubMed
    1. Cimadamore A, et al. Added clinical value of whole-mount histopathology of radical prostatectomy specimens: A collaborative review. Eur. Urol. Oncol. 2021;4:558–569. doi: 10.1016/J.EUO.2020.08.003. - DOI - PubMed
    1. Montironi R, Beltran AL, Mazzucchelli R, Cheng L, Scarpelli M. Handling of radical prostatectomy specimens: Total embedding with large-format histology. Int. J. Breast Cancer. 2012;1–6:2012. doi: 10.1155/2012/932784. - DOI - PMC - PubMed
    1. Rusu M, et al. Registration of presurgical mri and histopathology images from radical prostatectomy via rapsodi. Med. Phys. 2020;47:4177–4188. doi: 10.1002/MP.14337. - DOI - PMC - PubMed
    1. Samaratunga H, et al. International society of urological pathology (isup) consensus conference on handling and staging of radical prostatectomy specimens. working group 1: specimen handling. Mod. Pathol. 2011;24:6–15. doi: 10.1038/MODPATHOL.2010.178. - DOI - PubMed