Toward Routine Use of 3D Histopathology as a Research Tool

Three-dimensional (3D) reconstruction and examination of tissue at microscopic resolution have significant potential to enhance the study of both normal and disease processes, particularly those involving structural changes or those in which the spatial relationship of disease features is important. Although other methods exist for studying tissue in 3D, using conventional histopathological features has significant advantages because it allows for conventional histopathological staining and interpretation techniques. Until now, its use has not been routine in research because of the technical difficulty in constructing 3D tissue models. We describe a novel system for 3D histological reconstruction, integrating whole-slide imaging (virtual slides), image serving, registration, and visualization into one user-friendly package. It produces high-resolution 3D reconstructions with minimal user interaction and can be used in a histopathological laboratory without input from computing specialists. It uses a novel method for slice-to-slice image registration using automatic registration algorithms custom designed for both virtual slides and histopathological images. This system has been applied to >300 separate 3D volumes from eight different tissue types, using a total of 5500 virtual slides comprising 1.45 TB of primary image data. Qualitative and quantitative metrics for the accuracy of 3D reconstruction are provided, with measured registration accuracy approaching 120 μm for a 1-cm piece of tissue. Both 3D tissue volumes and generated 3D models are presented for four demonstrator cases. Three-dimensional (3D) reconstruction and examination of tissue at microscopic resolution have significant potential to enhance the study of both normal and disease processes, particularly those involving structural changes or those in which the spatial relationship of disease features is important. Although other methods exist for studying tissue in 3D, using conventional histopathological features has significant advantages because it allows for conventional histopathological staining and interpretation techniques. Until now, its use has not been routine in research because of the technical difficulty in constructing 3D tissue models. We describe a novel system for 3D histological reconstruction, integrating whole-slide imaging (virtual slides), image serving, registration, and visualization into one user-friendly package. It produces high-resolution 3D reconstructions with minimal user interaction and can be used in a histopathological laboratory without input from computing specialists. It uses a novel method for slice-to-slice image registration using automatic registration algorithms custom designed for both virtual slides and histopathological images. This system has been applied to >300 separate 3D volumes from eight different tissue types, using a total of 5500 virtual slides comprising 1.45 TB of primary image data. Qualitative and quantitative metrics for the accuracy of 3D reconstruction are provided, with measured registration accuracy approaching 120 μm for a 1-cm piece of tissue. Both 3D tissue volumes and generated 3D models are presented for four demonstrator cases. The three-dimensional (3D) reconstruction and examination of tissue at microscopic resolution have significant potential to enhance the study of normal and disease processes, particularly those involving structural changes or those in which the spatial relationship of disease features is important. Its application to, or combination with, techniques, such as immunohistochemistry (IHC) or in situ hybridization, adds further value by allowing understanding of additional phenotypic or functional information. Initial applications have been concerned with investigating the anatomical features and microarchitecture of normal tissue,1Kaufman M.H. Brune R.M. Baldock R.A. Bard J.B.L. Davidson D. Computer-aided 3D reconstruction of serially sectioned mouse embryos: its use in integrating anatomical organization.Int J Dev Biol. 1997; 41: 223-233PubMed Google Scholar tumor invasion, growth factor expression, and localization of therapeutic targets in relation to microvasculature and studying gene expression (eg, in developing mouse embryos2Han L. van Hemert J.I. Baldock R.A. Automatically identifying and annotating mouse embryo gene expression patterns.Bioinformatics. 2011; 27: 1101-1107Crossref PubMed Scopus (12) Google Scholar and the developing human brain3Wang X. Lindsay S. Baldock R. From spatial-data to 3D models of the developing human brain.Methods. 2010; 50: 96-104Crossref PubMed Scopus (4) Google Scholar). Competing alternative techniques to 3D tissue reconstruction using individually stained serial sections, and alternative nondestructive 3D imaging techniques, include optical projection tomography,4Quintana L. Sharpe J. Optical projection tomography of vertebrate embryo development.Cold Spring Harb Protoc. 2011; 2011: 586-594PubMed Google Scholar 3D imaging with ultrasonography,5Prager R.W. Ijaz U.Z. Gee A.H. Treece G.M. Three-dimensional ultrasound imaging.Proc Inst Mech Eng H. 2010; 224: 193-223Crossref PubMed Scopus (103) Google Scholar microscopic magnetic resonance imaging, or X-ray microcomputed tomography, confocal laser scanning or multiphoton microscopy, and serial block face imaging (eg, episcopic fluorescence image capture and high-resolution episcopic microscopy).6Denk W. Horstmann H. Serial block-face scanning electron microscopy to reconstruct three-dimensional tissue nanostructure.PLoS Biol. 2004; 2: 329Crossref Scopus (1101) Google Scholar Although all are mature technologies, which are used in research practice, each have limitations, whereas using conventional histopathological characteristics has significant advantages because it allows for conventional histopathological staining and interpretation techniques. More conventional methods of 3D histopathological analysis use proved and simple laboratory techniques to study structure, function, and disease manifestations. Examples include the use of photomicrographs and customized automated desktop software,7Wentzensen N. Braumann U.D. Einenkel J. Horn L.C. von Knebel Doeberitz M. Löffler M. Kuska J.P. Combined serial section-based 3D reconstruction of cervical carcinoma invasion using H&E/p16INK4a/CD3 alternate staining.Cytometry A. 2007; 71: 327-333Crossref PubMed Scopus (16) Google Scholar, 8Braumann U.D. Kuska J.P. Einenkel J. Horn L.C. Loffler M. Hockel M. Three-dimensional reconstruction and quantification of cervical carcinoma invasion fronts from histological serial sections.IEEE Trans Med Imaging. 2005; 24: 1286-1307Crossref PubMed Scopus (48) Google Scholar in which a digital camera mounted on a light microscope captured images from serial sections of cervical carcinoma invasion. An extension of this is the large-image microscope array,9de Ryk J. Namati E. Reinhardt J.M. Piker C. Xu Y. Liu L. Hoffman E.A. McLennan G. A whole organ serial sectioning and imaging system for correlation of pathology to computer tomography.Pro Biomed Opt Imag. 2004; 5: 224-234Google Scholar whereby sectioning, imaging, and reconstruction were used to reconstruct whole organs. A further example of a fully integrated system of 3D reconstruction was previously described,10Onozato ML, Merren M, Yagi Y: Automated 3D-reconstruction of histological sections. Presented at the International Academy of Digital Pathology Conference, Quebec, Canada, 2011Google Scholar in which a tissue processing, sectioning, slide scanning, and reconstruction system was fully automated and integrated into one process. Although these are useful methods for 3D reconstructions, particularly using an integrated system, the images and 3D reconstruction are limited by low resolution. Other limiting factors for conventional 3D histopathological analysis include the time and difficultly associated with acquiring many images with a microscope instead of an automated whole slide scanner,11Al-Janabi S. Huisman A. Van Diest P.J. Digital pathology: current status and future perspectives.Histopathology. 2011; https://doi.org/10.1111/j.1365–2559.2011.03814.xCrossref PubMed Google Scholar, 12Wu M.L.C. Varga V.S. Kamaras V. Ficsor L. Tagscherer A. Tulassay Z. Molnar B. Three-dimensional virtual microscopy of colorectal biopsies.Arch Pathol Lab Med. 2005; 129: 507-510PubMed Google Scholar, 13Petrie I.A. Flynn A.A. Pedley R.B. Green A.J. El-Emir E. Dearling J.L. Boxer G.M. Boden R. Begent R.H. Spatial accuracy of 3D reconstructed radioluminographs of serial tissue sections and resultant absorbed dose estimates.Phys Med Biol. 2002; 47: 3651-3661Crossref PubMed Scopus (2) Google Scholar the absence of a fully integrated system for reconstruction,14Namati E. De Ryk J. Thiesse J. Towfic Z. Hoffman E. McLennan G. Large image microscope array for the compilation of multimodality whole organ image databases.Anat Rec (Hoboken). 2007; 290: 1377-1387Crossref PubMed Scopus (13) Google Scholar and the significant amount of time associated with manual input previously required to guide 3D reconstruction.13Petrie I.A. Flynn A.A. Pedley R.B. Green A.J. El-Emir E. Dearling J.L. Boxer G.M. Boden R. Begent R.H. Spatial accuracy of 3D reconstructed radioluminographs of serial tissue sections and resultant absorbed dose estimates.Phys Med Biol. 2002; 47: 3651-3661Crossref PubMed Scopus (2) Google Scholar In light of these limiting factors, we have developed novel 3D histopathological software, which uses automated virtual slide scanners to generate high-resolution digital images and produce 3D tissue reconstructions at a cellular resolution level. It can be used on any stained tissue section (eg, H&E, IHC, or special stains or chromogenic in situ hybridization). It is based on a general image-based registration algorithm, which is reasonably robust over a wide variety of data (ie, it is not tuned to a specific data type/application) and operates using an integrated system that requires minimal manual intervention once the slides are divided into sections, stained, or mounted. The virtual slide scanners digitize automatically, and the software communicates with the image-serving software, which aligns the images and produces visualization in one integrated package. It uses high-resolution registration and high-performance computing that takes advantage of parallel computing using the OpenMP library in C++ (Microsoft, Redmond, WA) to use all available cores (n = 8) of our server. It also uses a novel means of multilevel registration (described later), in which the user can manually select a region, zoom in, and reregister the area. It then uses data fusion techniques to visualize microanatomical and functional information in conjunction with the structural 3D reconstruction and novel data visualization techniques to allow researchers to explore the resulting large data sets. In developing this novel 3D histopathological software, we are seeking to address a clinical and research need for high-resolution 3D microscopy. Many fields, including tumor biology, embryology, and transgenic models, would benefit from correlation of structure and function in 3D, but no current technology can integrate tissue microarchitecture, cellular morphological characteristics, and function on large tissue samples. This system has been used to generate 300 separate 3D tissue volumes from eight different tissue types, using a total of 5500 virtual slides comprising 1.45 TB of primary image data. This article describes the application of this method in four case studies. The following data sets were prepared (all shown in Figure 1): data set A, 111 sections of an 18-day post-fertilization mouse embryo (stained with H&E) (Figure 1A); data set B, 70 sections from a human liver containing a deposit of metastatic colorectal carcinoma adjacent to a blood vessel (stained with H&E) (Figure 1B); data set C, 100 sections from a cirrhotic human liver infected with hepatitis C (stained with picrosirius red) (Figure 1C); and data set D, 130 sections of a single rat glomerulus (stained with solochrome cyanin/phloxin) (Figure 1D). All human tissue was surplus surgical tissue with local ethical approval. The CD1 mouse embryo was obtained from a 10-week-old time-mated female (mated with a stud of the same strain), and the rat glomerulus was imaged from the kidney of a 12-week-old male Sprague-Dawley rat. Both were sacrificed under Schedule 1 of the Animals (Scientific Procedures) Act, 1986, under Project License PPL 40/2917. Tissue sections for data sets A, B, and C were prepared using standard histological techniques, as follows. Tissue was formalin fixed for 2 to 3 days, processed in a Leica ASP 200 tissue processor (Leica Microsystems, UK; Milton Keynes, UK) for 48 hours, and paraffin embedded. Serial sections (5 μm thick) were cut using a standard microtome, and selected sections (eg, every seventh section for data set A, giving a gap of 35 μm between images; every fifth section for data set B, giving a gap of 25 μm between images; and every 20th section for data set C, giving a gap of 100 μm between images) were stained using H&E and picrosirius red for the hepatitis C liver. Glass slides were scanned using Aperio Technologies, Inc. (San Diego, CA), T2 and T3 scanners using ×20 (for both of the human liver cases) and ×40 (for the embryo and glomerulus) images with a resolution of and μm Tissue sections for data set were prepared using histological techniques, as follows. Tissue was formalin and in Serial sections μm thick) were cut using an and every single section was selected and stained using solochrome The slides were scanned using Aperio Technologies, T2 and T3 scanners using a ×40 with a images with a resolution of μm sets A, B, and C were then used as in and data set was used as in data sets were into our customized software and using a slice-to-slice image-based registration virtual slide section to the of the tissue because this section the tissue and allows for the of the registration was used as a Serial sections to this section were to the using a slice-to-slice image-based registration from the with images to The set of images were then to a 3D data The slice-to-slice image registration used was a method based on an extension of E. C. of and images using Trans PubMed Scopus Google Scholar This method allows and to be to by the images to are by the virtual slides by a factor of This is then used as input to a registration method that the image into a set of that are individually E. C. of and images using Trans PubMed Scopus Google Scholar A is using a robust method to a set of from the set of are from each one at each and one at the This method was applied at image and 3 zoom zoom and zoom with the of the application the This was all integrated into a user-friendly software that images from the image and of the images with a Qualitative analysis was on all data sets to the accuracy of each in quantitative analysis was on data sets of the and a was generated for each data in which the and be The image shown in Figure has a the current resolution of slide in are to large at the resolution. The registration accuracy virtual slides can be as large as a single (described This is the of a with many required to over the image is to in the of local tissue and imaging accuracy reconstructions, we have an which is integrated into our software, by which the user can select of the image to at using the as an (ie, using a with This is in Figure This method is in to the multilevel resolution registration method previously C, automatic registration for large microscopy images. Presented at the and Scholar but uses resolution for registration as to using a single high-resolution with features for The from the whole image registration is used as an for this registration, which is based on the method previously D. D. A image registration with application to 3D with in Scholar this sections in which the same are in sections are The generated is then used to generate a 3D and of each data This software is based on software and from the available at W. The An to 3D Scholar which 3D of data in and of both are integrated into the reconstruction The 3D software has by in conjunction with a by which the user can the and to zoom in on and of The each image individually the data set and manually the to The software we have developed of 3D using a variety of methods manual is a method we in which the user an example by with the All with an a of this example are this be used a N. Digital A and Scholar to using the selected as a and as the is the user can then of the using a resolution 3D Scopus Google Scholar and a from the W. The An to 3D to generate a 3D of can be on of the data as in and can be used to and different of the reconstructed volumes of one with features to conventional cut Qualitative for all data sets was by the reconstructed volumes in (Figure and these one from each image in the can be as between This was the of In the that was a registration (Figure the images were and manually using a manual which is integrated into our This that the manually images were and were (Figure was on data sets (Figure containing features (eg, and blood to the sectioning and these features were using Aperio software Technologies, of were using the registration using our software, and the between was using the standard (ie, images using the Scopus Google Scholar were as shown in Figure and serial sections were to accuracy for all of image registration and to between images sections The accuracy of this is limited by the accuracy of the manual and the in from one section to we to be in the of to 100 of cirrhotic liver tissue stained with H&E with used for registration accuracy of cirrhotic liver tissue stained with picrosirius red with a single blood vessel used for registration accuracy Large Figure image set and were as the mouse embryo set the human liver tissue with metastatic colorectal carcinoma set the human liver cirrhotic tissue with hepatitis C set and the rat glomerulus set The time it takes for an reconstruction, from at the to at the for each case is shown in of the time is time to and reconstruct the images; of the human time is for sectioning and a of time is and the 3D reconstruction (ie, registration has and the zoom and the application for 3D liver hepatitis C rat by human time time human time (5 is by scanning from the rat slides in and computer time is by in a The 3D reconstruction for all data sets are shown in Figure The 3D of Figure the mouse and information on the anatomical morphological and features of different The 3D visualization the anatomical features of the mouse including the and and applications include anatomical and expression data and virtual of transgenic models. The 3D of Figure tumor morphological features and adjacent to and into the metastatic this into tumor and its to and the of such The 3D visualization of the in the liver tissue with an metastatic colorectal carcinoma The 3D of Figure information on and relationship to cirrhotic It also the the into and structure in relation to and and The 3D visualization cirrhotic by and the can also be It the potential for of information on the of the disease and It also the on a The 3D of Figure the single rat glomerulus by and the and on of the It the of the glomerulus and information on cellular such as the and the The 3D visualization the and the the the and selected It also the at a resolution. can use the and to the of to in the 3D and different can be on or that each can be individually in the 3D Qualitative are in Figure which different the of separate reconstructions of data set A have been to the same registration in the of and Figure at particularly the in the blood vessel whereas Figure no registration and the blood vessel is a (Figure registration accuracy metrics that can be in the accuracy metrics for the H&E case (Figure an with an of In the accuracy metrics for the picrosirius red case (Figure an with a standard of The registration whole virtual is 1 the or is registration accuracy over sections 120 μm for a piece of tissue μm large (ie, with H&E and over for (5 registration accuracy over sections are as the in a with and over for (5 slice-to-slice registration accuracy over sections are as the in in a are as the are as the in We a novel means of 3D histological reconstruction that whole slide scanning, image serving, registration, and visualization into one integrated user-friendly package. It produces high-resolution 3D reconstructions with minimal user interaction to sectioning, and can be N. of the automatic sectioning system for light the to sections on the by using 2002; Scopus (2) Google Scholar but we not have to such We have also that this software is and robust because we have from a of different and using serial and sections. We that this system the for the use of 3D histopathological analysis as a routine research is a of uses for this of 3D reconstructions and 3D the of large 3D data which allows for the analysis of tissue microanatomical and morphological automated or of 3D volumes can be (eg, to or and staining in a 3D the to study structural and functional information (eg, to expression at the of a tumor or growth factor expression in developing it can the of novel in microscopic such as in microscopy (eg, of in and high-resolution imaging, are The of robust 3D tissue visualization allow the of these novel imaging techniques the and this enhance Other have used 3D reconstruction to of colorectal carcinoma liver van Three-dimensional reconstruction of carcinoma in 1997; PubMed Scopus Google Scholar or the of cervical N. Braumann U.D. Einenkel J. Horn L.C. von Knebel Doeberitz M. Löffler M. Kuska J.P. Combined serial section-based 3D reconstruction of cervical carcinoma invasion using H&E/p16INK4a/CD3 alternate staining.Cytometry A. 2007; 71: 327-333Crossref PubMed Scopus (16) Google Scholar, 8Braumann U.D. Kuska J.P. Einenkel J. Horn L.C. Loffler M. Hockel M. Three-dimensional reconstruction and quantification of cervical carcinoma invasion fronts from histological serial sections.IEEE Trans Med Imaging. 2005; 24: 1286-1307Crossref PubMed Scopus (48) Google but such are limited in and by the need to the of available and software, and the of high-resolution images that can be reconstructed at any Although M.L.C. Varga V.S. Kamaras V. Ficsor L. Tagscherer A. Tulassay Z. Molnar B. Three-dimensional virtual microscopy of colorectal biopsies.Arch Pathol Lab Med. 2005; 129: 507-510PubMed Google Scholar have also used virtual slides to the of image registration using the high-resolution whole slide image has not been techniques of 3D reconstruction have also been developed but are Optical projection L. Sharpe J. Optical projection tomography of vertebrate embryo development.Cold Spring Harb Protoc. 2011; 2011: 586-594PubMed Google Scholar reconstruction the but is limited in resolution and to 3D imaging with ultrasonography,5Prager R.W. Ijaz U.Z. Gee A.H. Treece G.M. Three-dimensional ultrasound imaging.Proc Inst Mech Eng H. 2010; 224: 193-223Crossref PubMed Scopus (103) Google Scholar magnetic resonance imaging, or X-ray microcomputed tomography is nondestructive but is also limited by the resolution and a of functional information. laser scanning microscopy and multiphoton microscopy high-resolution 3D reconstruction, but are be Serial block face imaging fluorescence image capture and high-resolution episcopic W. Horstmann H. Serial block-face scanning electron microscopy to reconstruct three-dimensional tissue nanostructure.PLoS Biol. 2004; 2: 329Crossref Scopus (1101) Google Scholar resolution over of but is limited because it is not to image a the block has been and staining can be tissue use as previously N. of the automatic sectioning system for light the to sections on the by using 2002; Scopus (2) Google Scholar which address the in our 3D reconstruction which is the time to serial or sections on the use of a fully automated staining further the amount of manual input in this process. on developing whole slide 3D visualization at resolution. Although and reconstruction be at the resolution of the 3D visualization of the reconstruction is not at this resolution because of and resolution (eg, 3D obtained from this study were limited to images of be In future we to a system for of such data sets at all This in a 3D microscope for 3D volumes that would high-resolution such as a D. N. J. J. R.A. conventional microscope for pathology an PubMed Scopus Google Scholar for We and of Digital Pathology for scanning and computing of the article of 3D as a of the of The of the have the was a of in by the and

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.