| Abstract|| |
The last two decades have seen considerable progress in the use of digital technology in histopathology. Digital photography of microscopic slides and the use of static images gave way to robotic microscopes. These technologies had their own limitations that precluded their widespread use. Creation of whole slide scanners that can produce digitized whole slide images (WSI) and the “comparable to conventional microscope” experience opened multiple avenues for their utilization not only in specific applications such as expert consults, quality assessment programs, education and archiving, but also for routine day-to-day reporting. Industry pressures driven by consumer requirements have led to great development in image quality, speed of scanning, size of stored files, and capital cost of scanners. User-friendly software and analytical algorithms have further enhanced user experience. Challenges that need to be either accepted or overcome would include capital expense not significantly yielding a return on investment, and management of storage space. This review attempts to take the reader through the evolution of WSI scanners and to share the author's experience with WSI for routine histopathology reporting, education, and external quality assessment along with a review of available literature.
Keywords: Digital pathology, slide scanning, virtual microscopy, whole-slide imaging
|How to cite this article:|
Iyengar JN. Whole slide imaging: The futurescape of histopathology. Indian J Pathol Microbiol 2021;64:8-13
| Introduction|| |
What is “digital” in pathology practice? “It is an image-based information environment which is enabled by computer technology that allows for management of information generated from a digital slide” (Wikipedia). The usage of barcode labels on test request forms, worksheets, tissue processing cassettes, and microscopic slides by themselves could be considered as contributors to this digital environment. Microscopic examination of stained sections has been the only available instrumentation in surgical pathology for centuries. Static image capture through the use of conventional cameras and digital cameras has enabled the conversion of microscopic images into photographs and projection transparencies. Introduction of video cameras mounted on microscopes opened new ways of discussing microscopic images through local closed-circuit television screens and made the practice of “telepathology” possible.,
The first digital scanners were introduced around 2001 which helped combine the advantages of images from live cameras (whole-slide access) and digital cameras (high resolution). This technology has been undergoing rapid evolution through improved computer processing power, data transfer speeds, software, and cloud storage, which has resulted in the increasing use of digital pathology for activities such as archiving, telepathology, and image analysis. whole slide images (WSI) also known as “virtual microscopy” attempts to replicate what we achieve through light microscopy using computer-generated processes. The first process uses specialized hardware (scanners) to “digitize” glass slides to generate a large representative digital image. The second process employs virtual slide viewing software to view and analyze the digital files., Using modern WSI systems, the pathologist can navigate a virtual slide just like they navigate Google Maps.
Through this review, the author attempts to give an overview of digital scanners, compare digital microscopy with conventional microscopy, and highlight its applications in pathology.
A wide range of commercially available WSI scanners have been developed over the last decade each designed to meet specific applications. The first WSI scanners that were introduced in the late 1990s were quite primitive compared with their contemporary counterparts. WSI technology was inspired by pioneering efforts to achieve high-resolution scanning of entire glass slides. The virtual microscope created by Ferreira et al. in 1997 had major limitations by way of time taken and by the fact that it scanned only a single extended field. The next major development was the advent of an automated high-speed system created by Interscope Technologies that could capture entire slides at high resolution in a time-efficient fashion. Most modern WSI scanners are capable of producing high-resolution digital slides in the span of minutes. All WSI scanners used for virtual microscopy are made up of two essential components: the hardware and the software.
The WSI scanner is essentially a microscope under robotic control. The basic components of any type of WSI scanner include a) microscope with lens objectives, b) light source (bright field/fluorescence), c) robotics to load and to move the slides around, d) digital camera(s) that capture the image, e) computer, and f) software to manipulate, manage, and view the digital slides. Use of dynamic prefocusing functionality helps speed up the scanning process. Few scanners, especially those used in hematology, offer both dry and oil-immersion scanning (Aperio, CellaVision). The robotics in WSI scanners are the key to avoid breaking of slides, for stage accuracy (ability to localize specific parts of the section based on commands given), and for effective switching of objectives. They are also capable of traversing glass slides at speeds above 180 mm/s. A range of slide loading capacities are available ranging from single slide to small slide batches to practically hundreds of slides with continuous loading. Cost of the scanners is commensurate with these capabilities. The type of slides capable of being scanned range from dry slides (with coverslip) to wet slides and slides without a cover slip. Most scanners are equipped with bar-code readers and can read one or two dimensional bar codes. The average scanning speed varies from less than a minute to around 3 minutes per slide. Scanning times beyond this are not suitable for routine use. Z-stacking (ability to bring deeper parts of the section onto focus on the same plane to enhance depth of image) either is built into the process or can be enabled by choice. The area of slide to be scanned can also be defined by the user. Another important component in the process is the use of specialized imaging software that stitches the digital data together to produce a WSI that is free from overlaps and lost frames. This is achieved either by means of a tile-based scanning or a line-based scanning. Other high throughput scanners use advanced technology such as array microscopes and independent dual sensor scanning. Resolution of the scanned image depends on the objective used including its numerical aperture, and the quality of camera's photosensors. File size is directly proportional to the resolution and an average file scanned at 40× magnification may run into several gigabytes, though most modern scanners have very efficient file compression capabilities. This calls for increased storage capabilities to be made available in labs that use WSI in regular practice. Cloud-based storage is also an option that is being increasingly used.
In the author's personal experience of over 60 months of primary tissue diagnosis using WSI on the Roche Ventana platform with Virtuoso v. 5.6.2 software, the average file size of a single slide scanned at a magnification of 20× is 800 MB. File storage management is done in the following manner. Storage of the files is done in three stages: a) recent data of past 7 days on a hard disk of 4 TB capacity, b) remote but readily accessible data storage on a 50 TB hard disk (previous 18 months), and c) offline but retrievable archival of historic data currently on a 100 TB hard disk with the capability of adding further storage space.
This facilitates image acquisition, viewing, image management, sharing, and analysis. Some software packages include scoring and image analysis algorithms and other artificial intelligence tools. The software used to navigate digital slides—the image viewer, allows users to view and navigate virtual slides on a digital screen and is designed to provide an experience that is as close the microscopy as possible. Most viewers allow the users to zoom-in to magnifications a step higher than the one at which the slide was scanned. Zoom-out is achieved in a manner that is not possible with the conventional microscope. The user can experience a whole-slide view of the section [Figure 1]. Other software functions provided by different vendors include slide annotation tools (ability to mark/measure specific parts of section and to add notes about the findings), synchronized viewing of multiple sections, slide sharing, report formatting and so on. WSI software tools offer an advantage of annotation and ease of measurements including WPOI-5 in oral cancers, distance of tumor from resected margins, assessing the CRM in colorectal cancers, measurement of tumor sizes, micrometastases in lymph nodes, and other mandatory measurements that would be a challenge to perform using traditional microscopy [Figure 2]. Identification and enumeration of mitoses is a challenge that the author experienced with WSI when compared with conventional microscopy. It was initially difficult to accurately identify mitoses, which improved with experience. One of the contributing factors for improvement was the emphasis put into good slide preparation. Enumeration of mitoses and expressing the mitotic index/10 HPF becomes a challenge since the standard requirements of objective and eye-piece in a microscope do not apply to the WSI. However, algorithms using artificial intelligence are now available that have overcome this difficulty. Larger microorganisms such as fungal hyphae and larger bacterial colonies can be easily identified when slides are scanned at the usual 20× magnification. Smaller organisms such as H. pylori may pose a challenge and would require rescanning at higher magnifications?. The adaptability of software to a variety of independent WSI reading systems and ease of integration of the WSI software to Laboratory Information Systems (LIS) are challenges that need special consideration when choosing a WSI system.
|Figure 1: Whole view of section (HE 1×) to highlight the advantage of WSI. (a) Fibroadenoma breast showing all margins in one field. (b) Myositis ossificans highlighting the “reverse zonation”|
Click here to view
|Figure 2: Annotation and morphometry are easy with WSI (HE 4×). (a) Demonstration of WPOI-5 in oral cancer. (b) Measurement of micrometastasis in breast carcinoma|
Click here to view
Farahani et al. have elaborated comparative features of as many as 11 WSI systems in their publication.
Clinical applications of WSI
Introduction of WSI found immediate utility in research, as a teaching tool and for expert consultation. Its use for day-to-day clinical reporting has been increasing in the recent past with regulatory compliances being awarded. There are certain mandator?y prescan requirements that need to be satisfied for a successful WSI experience.
- Slide should be of a standard thickness not exceeding 1.2 mm and should have a uniform refractive index
- Tissue should be well processed
- Section should be thin and staining uniform refractive index
- Mounting medium should be of right quantity and optically clear
- Coverslip (either glass or film).
Certain degree of compromise in any of the above that might be overcome by the pathologist's “gut feel” with conventional microscopy does not work well with WSI.
[Table 1] lists out the applications of WSI applicable to routine diagnostics.
|Table 1: Salient features to be looked for in a WSI system based on the application intended|
Click here to view
A. WSI for primary tissue diagnosis:
Several challenges including workflow integration, technological infrastructure, acceptance by pathologists, and process validation need to be overcome when implementing WSI for routine diagnosis. Significant investment with little short-term financial returns also needs consideration. While countries like Canada, Singapore, and others in Europe have progressed considerably in using WSI for primary tissue diagnosis, others such as USA are catching up after putting more strict regulations in place. Implementation of WSI for primary tissue diagnosis is a dynamic process that moves through three phases: the preimplementation phase that involves defining of goals, needs, and objectives, identifying required infrastructure and securing resources; the implementation phase involving acquisition and building of infrastructure, training of the technical team and pathologists, definition of workflow, preparation of SOPs, validation of systems, and execution of process; and lastly th?e postimplementation involving assessment of efficiency and expansion to new applications. The entire process requires close liaison between the technologist, the systems administrator, IT, and the pathologist. During the initial phases of implementation, one needs to address concerns of pathologists who must use the WSI system in routine practice. It is important to engage pathologists early in the implementation process and help overcome their initial discomfort. The source of pathologist's discomfort is manifold ranging from lack of training, fear of making diagnostic errors, having to spend more time on a case at least initially, fundamental mechanical and ergonomic differences from the microscope, etc. Some pathologists find it more comfortable to use devises that replicate the microscope stage such as the 6DOF navigator, instead of the mouse or track-pad to navigate slide
Making working with WSI easier for pathologists requires careful design and understanding of workflow. Addition of one more step: the “digitization of slides” to the existing workflow in truth increases the turn-around time of both slide preparation and slide reporting. The personal experience of the author and published literature emphasize this., It is also cited that one of the potential reasons for a delay in reporting on WSI is the pathologist is simply not aware that virtual slides have been assigned for reporting and are “sitting” in the computer. This can be overcome by bringing in a system where the LIS can trigger a message to the pathologist with this information. The broad outline of workflow integration applied in our institution is depicted in [Figure 3]. Validation of the process of primary reporting using WSI is variably practiced by different institutions.,, The College of American Pathologists Pathology and Laboratory Quality Center developed certain guidelines for WSI validation. The salient points of recommendation that were formulated following review of 23 published studies are reproduced in [Table 2]
|Table 2: Guidelines for validating WSI systems for diagnostic purposes in Pathology (partly adapted from Pantanowitz et al. Arch Pathol Lab Med. 2013 Dec; 137 (12):1710-22.)|
Click here to view
In the personal experience of the author wherein 685 cases were analyzed for intraobserver variability, 18 cases showed discordance in the WSI report compared with the glass slide. A concordance rate of 98% was obtained. An interobserver variability study was also conducted with a combination of freshly recruited pathologists and pathologists with at least one year experience with WSI at our institution. A total of 246 sequential biopsy specimens were blinded and circulated to the pathologists along with all the clinical details available and the gross descriptions for each case. A washout period of at least 6 months was ensured. Findings are elaborated in [Figure 4]. It is obvious from this study that interobserver concordance improves with the experience of the pathologist
|Figure 4: Pie charts to highlight that interobserver concordance with WSI gets better with experience of the pathologist (a) Combination of “fresh” and experienced pathologists (b) only experienced pathologists|
Click here to view
There are certain scenarios where the pathologist may prefer to defer a WSI diagnosis to glass slide review. Few of these include difficult or unusual cases, slow digital workflow with need to release a report early, suboptimal image quality, suboptimal section quality, counting mitoses, and identification of microorganisms. The author's personal experience also concurs with this, and it is also understood that few of these situations may be overcome with experience. During the author's initial experience, categorization of lymphomas was more difficult with WSI when compared with glass slides. However, this was mitigated to a large extent with experience as well as by ensuring significant improvement in section and staining quality. This reiterates the need for the pathologist to develop that familiar “gut feel” even while working with WSI. Certain process adjustments are also warranted including the format of the label. This may vary from equipment to equipment. We modified our label format to comply with the CAP guidelines for uniformity of labelling of histopathology slides
WSI has a positive impact on TAT in network laboratories that receive slides/tissues from multiple peripheral centers for either primary diagnosis or a review report. Installation of a low-end slide scanner (depending on sample load) and setting up a tissue processing facility in these centers enables histopathology reports as well as frozen section reports to be released with a shortened turn-around time. This “hub and spoke” model, however, will require to be validated and the quality of slide preparation will need to be constantly monitored
B) WSI for Expert Consults: The obvious advantage of WSI for expert consults lies in the obviation of need to transport precious glass slides across cities/countries. Time saving and prevention of loss/breakag?e are major advantages. Challenges include adaptability of the expert to digital reporting, ability of software to allow for proper and accurate communication between the referring lab and the expert, ability of the expert to access clinical and other background data through the slide viewing software, and compatibility of slide viewing software across various platforms
c) WSI as a teaching/training tool: WSI has been used with success for numerous educational activities including multidisciplinary graduate and professional education, virtual tracking, performance improvement programs, and even for medical examinations. It has also found place as illustrations in journals and textbooks. Virtual slides offer several advantages over conventional glass slides by widening the scope of material that can be shared and not having to limit oneself to large tumors that can be recut onto multiple glass slides. The WSI do not fade over time and can be easily archived and retrieved when required. The author has had personal experience in conducting multiple teaching sessions and slides seminars on a nation-wide scale using WSI, with great degree of adaptability and acceptance from the participant
D) WSI for proficiency testing programs: Most providers of PT programs for histopathology have either in part or completely migrated to the WSI platform. This adds the obvious advantage of not having to restrict to recut of large tissues to provide a glass slide to each of the participants. It also enables simultaneous viewing by an infinite number of participants without compromising the image quality. Use of glass slides limits the choice of tissue, limits the number of participants, and at times forces the PT provider to adopt to a sequential viewing of slides by the participants. This in turn increases the time taken to complete each cycle. The author has been involved in two nationwide PT programs for histopathology over the past 14 years beginning with the use of glass slides and then migrating to WSI (www.ilqabangalore.com which is now closed and www.neu-qap.com). The transition from glass slides to WSI began in a phased manner in March 2012 and a complete switch was made to WSI in October 2015. Barring objections from few senior pathologists, the move was accepted by most participants (unpublished data).
Artificial intelligence in WSI reporting
Artificial intelligence (AI) is dubbed to be the “Third Revolution” in pathology, coming on the back of digital pathology. Evolution of digital pathology particularly WSI has created the perfect launching pad for the development of diagnostic and predictive algorithms using machine deep learning and other artificial intelligence tools. Algorithms for scoring of IHC of breast tumor markers, Pdl-1 reporting, Gleason scoring on H and E sections of prostate, localizing and accurately counting mitoses, and quantification of fibrosis are only a few examples., The translation of AI into clinical practice will require applications to be embedded seamlessly within digital pathology workflows, driving an integral approach to diagnostics, and providing pathologists with new tools that accelerate workflow and improve diagnostic consistency and reduce errors.
The Indian scenario
Digital pathology has made significant inroads into pathology teaching and diagnosi?s albeit in institutions and larger laboratories. The future will see greater presence of this technology in medium to small laboratories. This will be aided by increasing awareness, greater comfort levels, and not to mention, reduction in cost of scanners and digital storage options.,
SUMMARY AND CONCLUSION
Digital patholo?gy—WSI, the first revolution in modern pathology, has opened limitless options for diagnosis, institutional cross-consults, teaching, archival, and proficiency testing activities. It has provided a platform on which artificial intelligence algorithms can be developed through the machine and deep learning protocols. The contribution of this technology to precision medicine is being increasingly recognized. We pathologists are at the crossroads and need to take the call to take this journey or be left behind.
All colleagues of department of Histopathology for participating in the validation studies.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
| References|| |
Hanna MG, Pantanowitz L. Bar coding and tracking in pathology. Surg Pathol 2015;8:123-35.
Kayser K, Kayser G, Radziszowski D, Oehmann A. New developments in digital pathology: From telepathology to virtual pathology laboratory. Stud Health Technol Inform 2004;105:61-9.
Weinstein RS. Prospects for telepathology. Hum Pathol 1986;17:433-4.
Gilbertson JR, Patel AA,,Yagi Y. Clinical slide digitization–whole slide imaging in clinical practice. In: Gu J, Olgilive RW, et al
, editors. Virtual Microscopy and Virtual Slide in Teaching, Diagnosis, and Research. Boca Rotan, FL: Taylor and Francis; 2005.
Gifford AJ, Colebatch AJ, Litkouhi S, Hersch F, Warzecha W, Snook K, et al
. Remote frozen section examination of breast sentinel lymph nodes by telepathology. ANZ J Surg 2012;82:803-8.
Leong FJ, Leong AS. Digital photography in anatomical pathology. J Postgrad Med 2004;50:62-9.
] [Full text]
Huisman A, Looijen A, van den Brink SM, van Diest PJ. Creation of a fully digital pathology slide archive by high-volume tissue slide scanning. Hum Pathol 2010;41:751-7.
Zarella MD, Bowman D, Aeffner F, Farahani N, Xthona A, Absar SF, et al
. A practical guide to whole slide imaging: A white paper from the digital pathology association. Arch Pathol Lab Med 2019;143:222-34.
Pantanowitz L, Evans A, Pfeifer J, Collins L, Valenstein P, Kaplan K, et al
. Review of the current state of whole slide imaging in pathology. J Pathol Inform 2011;2:36-50.
] [Full text]
Ferreira R, Moon B, Humphries J, Sussman A, Saltz J, Miller R, et al
. The Virtual Microscope. Proceedings : a conference of the American Medical Informatics Association / ... AMIA Annual Fall Symposium. AMIA Fall Symposium. 1997;449-453.
Farahani N, Parwani AV, Pantanowitz L. Whole slide imaging in pathology: Advantages, limitations, and emerging perspectives. Pathol Lab Med Int 2015;2015:21-33.
Weinstein RS, Descour MR, Liang C, Barker G, Scott KM, Richter L, et al
. An array microscope for ultrarapid virtual slide processing and telepathology. Design, fabrication, and validation study. Hum Pathol 2004;35:1303-14.
Stathonikos N, Veta M, Huisman A, van Diest P. Going fully digital: Perspective of a Dutch academic pathology lab. J Pathol Inform 2013;1:15.
Baidoshvili A, Bucur A, van Leeuwen J, van der Laak J, Kluin P, van Diest PJ. Evaluating the benefits of digital pathology implementation: Time savings in laboratory logistics. Histopathology 2018;73:784-94.
Têtu B, Evans A. Canadian licensure for the use of digital pathology for routine diagnoses: One more step toward a new era of pathology practice without borders. Arch Pathol Lab Med 2014;138:302-4.
Evans AJ, Salama ME, Henricks WH, Pantanowitz L. Implementation of whole slide imaging for clinical purposes: Issues to consider from the perspective of early adopters. Arch Pathol Lab Med 2017;141:944-59.
Molin J, Lundström C, Fjeld M. A comparative study of input devices for digital slide navigation. J Pathol Inform 2015;6:7.
] [Full text]
Hanna MG, Reuter VE, Hameed MR, Tan LK, Chiang S, Sigel C, et al
. Whole slide imaging equivalency and efficiency study: Experience at a large academic center. Mod Pathol 2019;32:916-28.
Bauer TW, Slaw RJ. Validating whole-slide imaging for consultation diagnoses in surgical pathology. Arch Pathol Lab Med 2014;138:1459-65.
Goacher E, Randell R, Williams B, Treanor D. The diagnostic concordance of whole slide imaging and light microscopy: A systematic review. Arch Pathol Lab Med 2017;141:151-61.
Snead DRJ, Tsang Y-W, Meskiri A, Kimani PK, Crossman R, Rajpoot NM, et al
. Validation of digital pathology imaging for primary histopathological diagnosis. Histopathology 2016;68:1063-72.
Pantanowitz L, Sinard JH, Henricks WH, Fatheree LA, Carter AB, Contis L, et al
. Validating whole slide imaging for diagnostic purposes in pathology: Guideline from the college of American pathologists pathology and laboratory quality center. Arch Pathol Lab Med 2013;137:1710-22.
Brown RW, Speranza VD, Alvarez JO, Eisen RN, Frishberg DP, Rosai J, et al
. Uniform labeling of blocks and slides in surgical pathology: Guideline from the college of American pathologists pathology and laboratory quality center and the National society for histotechnology. Arch Pathol Lab Med 2015;139:1515-24.
Chong T, Palma-Diaz MF, Fisher C, Gui D, Ostrzega N, Sempa G, et al
. The California telepathology service: UCLA's experience in deploying a regional digital pathology subspecialty consultation network. J Pathol Inform 2019;10:31.
] [Full text]
Jones NC, Nazarian RM, Duncan LM, Kamionek M, Lauwers GY, Tambouret RH, et al
. Interinstitutional whole slide imaging teleconsultation service development: Assessment using internal training and clinical consultation cases. Arch Pathol Lab Med 2015;139:627-35.
Pantanowitz L, Szymas J, Wilbur D, Yagi Y. Whole slide imaging for educational purposes. J Pathol Inform 2012;3:46.
] [Full text]
Salto-Tellez M, Maxwell P, Hamilton P. Artificial intelligence-The third revolution in pathology. Histopathology 2019;74:372-376.
Parwani AV. Next generation diagnostic pathology: Use of digital pathology and artificial intelligence tools to augment a pathological diagnosis. Diagn Pathol 2019;14:138.
Niazi MKK, Parwani AV, Gurcan MN. Digital pathology and artificial intelligence. Lancet Oncol 2019;20:253-61.
Ryu HS, Jin M-S, Park JH, Lee S, Cho J, Oh S, et al
. Automated gleason scoring and tumor quantification in prostate core needle biopsy images using deep neural networks and its comparison with pathologist-based assessment. Cancers 2019;11:1860.
Serag A, Ion-Margineanu A, Qureshi H, McMillan R, Saint Martin M-J, Diamond J, et al
. Translational AI and deep learning in diagnostic pathology. Front Med 2019;6:185.
Chandra M. Digital pathology slides in medical education. Indian J Dermatopathol Diagn Dermatol 2014;1:17. [Full text]
Chordia TD, Vikey A, Choudhary AB, Samdariya Y, Chordia DS. Current status and future trends in telepathology and digital pathology. J Oral Maxillofac Pathol 2016;20:178-82.
] [Full text]
Jayaram N Iyengar
Department of Histopathology, Anand Diagnostic Laboratory (A Neuberg Associate), Anand Tower, 54 Bowring Hospital Road, Shivajinagar, Bangalore - 560 001, Karnataka
Source of Support: None, Conflict of Interest: None
[Figure 1], [Figure 2], [Figure 3], [Figure 4]
[Table 1], [Table 2]