Let us trace our path from glass slide to virtual slide and artificial intelligence.

Abstract

Beginning in the late 1980s, we developed methods for digital image acquisition, virtual microscopy, and quantitative image analysis that anticipated many features of contemporary digital pathology. Our work included machine vision systems for the automated detection and characterization of high-grade prostatic intraepithelial neoplasia, malignancy-associated changes, and prostate cancer, as well as quantitative approaches for identifying cribriform architecture and early decision support systems for diagnostic pathology. These pioneering studies established methodological foundations that have evolved into modern whole-slide imaging, multiplex tissue analysis, and AI-based diagnostic algorithms. Recent advances in deep learning have further expanded these concepts, improving cancer detection, grading, and prognostic assessment. Although AI has considerable potential to enhance the accuracy and efficiency of prostate pathology, its current role is best viewed as an assistive tool integrated with expert pathological interpretation.