The SPIE 2025 Digital and Computational Pathology conference highlighted several recent advancements in the field:

• Novel computational approaches using graph theory and AI identified distinct patterns in rapid versus classic Alzheimer’s disease progression, offering new insights into disease heterogeneity.
• An AI system was developed to analyze immunohistochemistry slides for TP53 mutations in AML, potentially accelerating testing from weeks to minutes.
• A new AI framework achieved high accuracy in predicting prostate cancer outcomes without relying on traditional Gleason grading, demonstrating the potential to move beyond conventional diagnostic approaches.
• Advanced AI pipelines enhanced tissue quantification in cerebral aneurysm analysis and cell type prediction from H&E slides, promising to streamline histopathology workflows.
• New approaches were presented to make AI models more interpretable and trustworthy, including correlating AI-derived features with traditional pathological markers and developing human-in-the-loop frameworks.
• Multiplexed imaging and computational methods revealed distinct cellular neighborhoods in quiescent Crohn’s disease tissue, challenging assumptions about disease remission.

Close collaboration between pathologists, computer scientists, and other stakeholders will be crucial for shifting digital pathology from pattern recognition to scientific discovery and multimodal integration across the patient’s diagnostic pathway. Key areas of focus include:

• Developing more flexible and adaptive AI systems that can learn and evolve like clinical teams.
• Integrating information across multiple modalities and the entire patient journey.
• Addressing implementation challenges around digital workflows and reimbursement.
• Ensuring responsible development and use of AI in healthcare.

We hope you’ve enjoyed this series. To find individual episodes, please visit the Conference Reports page on Pathology News.