Digital and Computational Pathology: What a Time to Be Alive!

Abstract

There is no doubt that the field of pathology is undergoing radical transformation, thanks to its digitalization and the emergence of artificial intelligence (AI). What was once a discipline that was heavily reliant on manual examination and interpretation under the microscope is rapidly entering an era where vast amounts of data will increasingly be managed in entirely digital workflows and analyzed with unprecedented precision and efficiency. The incorporation of AI and other digital tools is reshaping the specialty in a way that is not merely an incremental change, but a revolutionary leap forward. The rise of fully digital laboratory environments, coupled with a growing array of AI-driven diagnostic tools, is set to streamline tasks for pathologists by expediting turnaround times, minimizing errors, and offering a comprehensive view of patient data. These advancements promise to boost the efficiency of pathology workflows and enhance patient care quality by providing more quantitative, accurate, and consistent diagnoses.1Niazi M.K.K. Parwani A.V. Gurcan M.N. Digital pathology and artificial intelligence.Lancet Oncol. 2019; 20: e253-e261https://doi.org/10.1016/S1470-2045(19)30154-8Abstract Full Text Full Text PDF PubMed Scopus (534) Google Scholar Furthermore, the incursion of digitalization and AI into pathology is set to reshape the very structure of pathology departments and their workforce dynamics, bestowing new roles upon both pathologists and technicians. In the vast tapestry of medical evolution, it is a rare spectacle to witness a discipline undergo a transformation as fast and as profound as the one we are currently seeing with pathology. This rapid metamorphosis can be attributed to a synchrony of technological advances: the emergence of new hardware tools, such as digital slide scanners, alongside improvements in software, specifically neural networks that are specialized in analysis and interpretation of imaging data. Coupled with enhanced solutions for efficient storage and management of whole-slide images (WSI), these advancements are almost simultaneous and push the discipline into uncharted territories. The timely article by Qureshi et al2Qureshi HA, Chetty R, Kuklyte J, et al. Synergies and challenges in the preclinical and clinical implementation of pathology AI applications. Mayo Clinic Proceedings: Digital Health. Published online 2023.Google Scholar published in this issue of Mayo Clinic Proceedings: Digital Health offers a valuable compass for navigating this journey. It combs through the literature, casting light on areas that are primed for an early adoption of Digital and Computational Pathology (DCP), while also highlighting the most pressing challenges on the horizon. After reviewing the current landscape of digital slide scanners and WSI viewers and formats, the authors delve into the recent literature and identify the hottest areas of research in DCP, which suggest a potential for an earlier adoption in clinical settings. Among these, their work discusses applications such as case prioritization and abnormality detection for guiding pathologists' focus to more urgent cases and crucial image regions, quality control of scanned WSIs, detection and quantification of cells and prognostic biomarkers, or content-based image retrieval,3Sridhar A. Doyle S. Madabhushi A. Content-based image retrieval of digitized histopathology in boosted spectrally embedded spaces.J Pathol Inform. 2015; 6: 41https://doi.org/10.4103/2153-3539.159441Crossref PubMed Google Scholar ie, retrieval of similar cases for a particular query study, with potential to provide statistical data as well as prognostic information. Yet, intertwined with these promising arenas are inherent challenges that the field must face. The dearth of algorithms currently available in regular clinical use makes it very difficult to accurately evaluate their real impact in a hospital setting.4Kearney S.J. Lowe A. Lennerz J.K. et al.Bridging the gap: The critical role of regulatory affairs and clinical affairs in the total product life cycle of pathology imaging devices and software.Front Med (Lausanne). 2021; : 8https://doi.org/10.3389/fmed.2021.765385Crossref Scopus (6) Google Scholar Nonetheless, ethical concerns regarding the possibility of breaches in data privacy, inherent biases in algorithms, potential harm caused by incorrect AI-generated results, aggravation of disparities in healthcare, and possibility of AI use leading to deskilling of pathologists have been raised.5Chauhan C. Gullapalli R.R. Ethics of AI in pathology.Am J Pathol. 2021; 191: 1673-1683https://doi.org/10.1016/j.ajpath.2021.06.011Abstract Full Text Full Text PDF PubMed Scopus (22) Google Scholar Qureshi et al6Müller H. Holzinger A. Plass M. et al.Explainability and causability for artificial intelligence-supported medical image analysis in the context of the European In Vitro Diagnostic Regulation.New Biotechnol. 2022; 70: 67-72https://doi.org/10.1016/j.nbt.2022.05.002Crossref PubMed Scopus (23) Google Scholar further comment on issues such as ensuring reproducibility of AI outcomes across different labs, addressing cross-scanner discrepancies and variations in display, common hurdles in digitalization adoption which prominently include staff training, and legal and regulatory implications of computational pathology tools, such as those involved in explainability of AI outcomes. Moreover, it is essential to address the economic implications of the implementation of AI in hospital settings. The process encompasses not only the deployment of the actual algorithms, but also the potential acquisition of specific hardware and software, training of the staff operating the newly acquired technology, integration of the new tools with the already existing workflow, and additional costs related to system updates, maintenance, and monitoring. These considerations might exacerbate the existing disparities within the healthcare systems, as some hospitals will inevitably find it challenging to bear the considerable upfront costs of DCP implementation. Nonetheless, this investment can prove profitable, as shown in the analysis by Griffin and Treanor7Griffin J. Treanor D. Digital pathology in clinical use: where are we now and what is holding us back?.Histopathology. 2017; 70: 134-145https://doi.org/10.1111/his.12993Crossref PubMed Scopus (180) Google Scholar and as reported by the increased efficiency of results obtained at the Granada University Hospitals8Retamero J.A. Aneiros-Fernandez J. del Moral R.G. Complete digital pathology for routine histopathology diagnosis in a multicenter hospital network.Arch Pathol Lab Med. 2020; 144: 221-228https://doi.org/10.5858/arpa.2018-0541-OACrossref PubMed Scopus (89) Google Scholar and the Memorial Sloan Kettering Cancer Center9Hanna M.G. Reuter V.E. Samboy J. et al.Implementation of digital pathology offers clinical and operational increase in efficiency and cost savings.Arch Pathol Lab Med. 2019; 143: 1545-1555https://doi.org/10.5858/arpa.2018-0514-OACrossref PubMed Scopus (66) Google Scholar after implementing digital pathology. The potential benefits of widespread adoption of DCP are compelling, making it well worth the effort to overcome the associated challenges. The efficiencies of DCP promise to bridge the growing gap between the escalating demand for studies and the scarcity of pathologists, whereas the potential for more quantitative, precise, and reproducible diagnoses will definitely impact patient management for the better. Also, speeding up manual, repetitive tasks, such as cell counting, will ensure the optimal allocation of human expertise where it is most needed, allowing pathologists to participate in more valuable tasks, such as analysis of complex cases and engaging in multidisciplinary boards or research activities. Additionally, as highlighted by Qureshi et al, the quantitative capabilities of DCP might unveil new biomarkers in pathology images that remain imperceptible to the human eye. This concept, referred to as “pathomics,”10Colvin R.B. Getting out of flatland: Into the third dimension of microarrays.Am J Transplant. 2007; 7: 2650-2651https://doi.org/10.1111/j.1600-6143.2007.02024.xAbstract Full Text Full Text PDF PubMed Scopus (9) Google Scholar,11Solez K. Racusen L.C. The Banff classification revisited.Kidney Int. 2013; 83: 201-206https://doi.org/10.1038/ki.2012.395Abstract Full Text Full Text PDF PubMed Scopus (88) Google Scholar might hold significant potential for diagnostic and prognostic insights. Although all of this may sound like a vision from the distant future, it is not as far off as one might think. In fact, experts in the field believe that by 2030, this will be the day-to-day reality for most pathology departments.12Berbís M.A. McClintock D.S. Bychkov A. et al.Computational pathology in 2030: a Delphi study forecasting the role of AI in pathology within the next decade.EBiomedicine. 2023; 88104427https://doi.org/10.1016/j.ebiom.2022.104427Abstract Full Text Full Text PDF PubMed Scopus (17) Google Scholar In the midst of all these breakthroughs, one cannot help but echo the sentiment, "What a time to be alive!" The novelties poised to emerge in DCP in the near future are electrifying. This journal, committed to its mission to provide a platform for investigators, clinicians, and investigators to share in the disruptive transformation of medical practice through digital health advances,13Lopez-Jimenez F. Digital health in the 21st century: The best is yet to come.Mayo Clinic Proceedings: Digit Health. 2023; 1: 52-53https://doi.org/10.1016/j.mcpdig.2023.03.001Abstract Full Text Full Text PDF Google Scholar will undoubtedly be a leading platform for chronicling the advancements and breakthroughs within the realm of DCP. The current advancements are paving the way for this transformative change. In this regard, the article by Qureshi et al very nicely serves as a foundational touchstone, capturing the current state of DCP in the inaugural year of this Journal. Assuredly, this publication will remain at the forefront, closely monitoring the evolution of this discipline. What an era to witness, and indeed, what a time to be alive in the world of Digital and Computational Pathology. M. Álvaro Berbís is CEO and Board Member of Cells IA Technologies. The author thanks Dr. Reyes Sanles-Falagan for her help with manuscript formatting and editing. Synergies and Challenges in the Preclinical and Clinical Implementation of Pathology Artificial Intelligence ApplicationsMayo Clinic Proceedings: Digital HealthVol. 1Issue 4PreviewRecent introduction of digitalization in pathology has disrupted the field greatly with the potential to change the area immensely. Digital pathology has created the potential of applying advanced quantitative analysis and artificial intelligence (AI) to the domain. In this study, we present an overview of what pathology AI applications have the greatest potential of widespread adoption in the preclinical domain and subsequently, in the clinical setting. We also discuss the major challenges in AI adoption being faced by the field of digital and computational pathology. Full-Text PDF Open Access

Ähnliche Arbeiten