"The Final Adaptation of Routine AI"

The future of digital pathology is no longer a question of if, but how soon. AI has progressed from research to clinical tools. The remaining challenge is no longer technical…. It is a regulatory, operational, financial, organizational, and ultimately a human adaptation challenge. Pathology is at an inflection point. The essential building blocks are falling into place, and the remaining barriers are increasingly those of implementation science.

This presentation explores the final adaptation stages required to bring AI from successful pilot projects into routine pathology practice. Joe will tell 3 practical stories from academia, industry, and the advocacy field to outline practical conditions that enable sustainable clinical implementation. Topics include evolving regulatory frameworks, validation and quality management, emerging reimbursement and payment models, laboratory operations and validation, workforce transformation, and the emerging role of pathologists as leaders of AI-enabled diagnostic systems. Laboratories that successfully navigate these final adaptations will be well positioned to realize the promise of digital pathology and AI while improving patient care. 

“Agentic AI in the pathology lab: From hype to practical implementation”

Agentic AI has generated significant excitement for its potential to transform pathology workflows, but its practical implementation in the laboratory requires careful consideration of real-world constraints, governance, safety, validation, and integration with existing clinical systems. This presentation will explore the opportunities and limitations of agentic AI in pathology, with a focus on practical use cases, implementation challenges, and strategies for responsible deployment. The talk will highlight where agentic AI may add value, where caution is warranted, and how pathology teams can prepare for the evolving role of AI-enabled systems in the digital lab.

"Piña Coladas and Pixels: Building a Scalable Digital Pathology and AI Ecosystem"

Introduction
Digital pathology is rapidly evolving from isolated whole-slide imaging applications into an increasingly complex ecosystems incorporating image management systems (IMS), artificial intelligence (AI), laboratory information systems (LIS), and multiple diagnostic workflows. Successfully scaling this technology demands an infrastructure capable of integrating new technologies while maintaining efficient, reliable, and clinically meaningful workflows. CorePlus, a pathology laboratory based in Puerto Rico, has undertaken a progressive digital transformation aimed at developing such an ecosystem across both histology and cytology.

Methods
A phased approach was used to implement and expand digital pathology across the laboratory. Following implementation and validation of whole-slide imaging for routine histopathology, AI applications were progressively introduced to support the evaluation of prostate and breast specimens. Digital cytology workflows subsequently extended these capabilities to gynecologic and urine cytology samples. In collaboration with Syndeo, our affiliated information technology organization, CorePlus has developed the infrastructure required to support image acquisition, management, computational analysis, data movement, cybersecurity, storage, and clinical workflow across multiple platforms. Attention has been paid to interoperability and to reducing the technological and operational silos that can emerge as laboratories adopt solutions from multiple vendors. Rather than adopting separate digital solutions, the organization has focused on creating an integrated system that can grow and adapt over time.

Results
This approach resulted in an expanding digital pathology and AI ecosystem supporting multiple specimen types, diagnostic disciplines, and technology platforms. The progression from histology to cytology demonstrated that scalable deployment requires more than the validation and implementation of individual algorithms. Infrastructure, workflow design, interoperability, information technology support, and the ability to accommodate technologies with different requirements proved essential to expansion. Rather than maintaining independent digital solutions, the laboratory has progressively moved toward an integrated environment designed to grow and adapt as new technologies are introduced. Current development is focused on deeper integration with CorePlus’ internally developedLIS, with the goal of bringing clinical information, digital slides, AI results, and pathologist workflows together at the point of diagnosis.

Conclusion
The CorePlus experience demonstrates that enterprise-scale digital pathology and AI deployment is an evolutionary process requiring coordinated technical, operational, validation, and clinical strategies. Experience across prostate and breast histopathology and gynecologic and urine cytology highlights the importance of designing digital infrastructure for scalability and interoperability from the outset. Sustainable implementation should focus on building an ecosystem capable of incorporating today’s as well as tomorrow’s technologies.