| Abstract: |
The advancement of technology in the last several years has opened new ways to transform the practice of pathology from using glass slides, which are viewed under a microscope, to studying the microscopic histology as digital images to render a pathological diagnosis. This is an opportune time for pathologists to use their advanced pattern recognition capability to take it several steps further to remove subjective and qualitative aspects of diagnostic pathology such as mitoses counting, necrosis assessment, biomarker estimation to an objective, quantitative and more accurate analysis, allow for better workflow and reduce time-constrained activities like screening for metastases in lymph nodes, and detecting foci of cancer in a background of benign tissue. An additional benefit is remote review of histological images from sites other than designated Clinical Laboratory Improvement Amendments-approved facilities, with instantaneous availability of current and archival material allowing for remote consultations leading to rapid and smooth clinical care. An exciting area of rapid growth is the development of machine learning and artificial intelligence-based algorithms that can decipher the vast amount of data buried in histopathological images and thus unraveling molecular level changes such as detection of genetic mutations, protein expression, microsatellite instability, and new biomarker discovery. As these algorithms improve with the increased availability of large datasets, the transparency and interpretability of the algorithms will improve, leading to better standardization and interoperability, with regulatory and ethical guidelines for effective use in the clinical arena. These advances will ultimately improve patient diagnosis, clinical outcome and enhance the practice of pathology. © 2025 Elsevier Inc. All rights are reserved, including those for text and data mining, AI training, and similar technologies. |
