PV24 Speakers

Subject to change.

 

 

image

Bryan Dangott, MD

Director of Clinical Informatics, Mayo Clinic Florida


Bryan Dangott, MD is the Director of Clinical Informatics and a Consultant in the Department of Laboratory Medicine and Pathology at Mayo Clinic Florida.  Current research interests include digital pathology adoption, optimizing workflow, operational analytics, image analysis, and AI development.  In prior roles he served as Vice Chair of Pathology Informatics and Analytics at East Carolina University and as Medical Director for Flagship Biosciences and Interpace Pharma Solutions. 

 

 

SESSIONS

Going Digital - Roadmap for Implementation
   Mon, Nov 4
   12:45PM - 01:05PM ET

Transitioning from analog to digital workflows with digital pathology involves many considerations which may seem daunting without developing the right plan, expectations, and support. Our department began converting to digital workflows in early 2021 and now has all pathologists reading digitally for greater than 90% of the case volume (excluding hematology and cytology). We will discuss the process and lessons learned for scanner selection, slide preparation, PACS, LIS integration, infrastructure, deployment, and workflow adoption.Optimal scanner selection is probably the most critical component for going digital. While there are various scanners on the market, careful consideration should be given to choosing the right scanner for the intended application. An assessment of image quality and magnification requirements should rank highly in the factors to consider. Scanner redundancy and robustness may also be high priorities for scanners deployed in time sensitive environments such as frozen section or other intra-procedural environments. Ultimately, multiple scanner platforms may be needed to accommodate various modalities such as paraffin sections, frozen sections, immunoflourescence, and smears from hematology or cytology samples. The daily slide volume, scanner throughput, and expected turn around time will impact the total number of scanners needed for deployment.As processes move from manual to automated workflows, slide preparation may be subject to tighter tolerances and requirements. While slide bar codes are common in routine histology, hand labeling methods used by many frozen section laboratories may not be compatible with automated case assembly. In addition, outside consult cases may need to be relabeled in a manner that is compatible with internal accessioning systems. Slide preparation steps such as cover slipping, staining, and defect detection may also impact scanning operations. For example, cover slips or label stickers which extend beyond the edge of the glass slide can impact automated slide handling equipment. Wet slides or adhesive residue can impact slide gripper mechanisms and contribute to scanner downtime. Air bubbles, fingerprints, and debris can cause out of focus regions or obscure histology and thus require re-scan. Tissue layout on the slide can impact scanner throughput or even lead to scanner errors or re-scans if the tissue is too close to the edge of the slide. Stain intensity can impact automated tissue detection algorithms and result in tissue being missed by the scanner.Pathologists using glass slides may have a variety of methods for sorting and working through flats of glass. The case management system, PACS, and slide viewer should provide a comfortable and efficient environment to transition some of those analog workflows to digital workflows. In many ways, the digital environment has an advantage because the capacity to sort, annotate, and navigate slides can all happen on one screen. However, if core functions are missing or poorly implemented, pathologists may consider these as defects and detractors from going digital. Familiarizing and training pathologists how to optimally use the software is essential. In addition, at the elbow support is helpful in preventing frustration. Finally, pathologist workstations may need to be upgraded to handle slide viewing and case management functions.Any digital pathology operation requires a plan for hosting and storing whole slide images. Some groups prefer to purge digital images a few months after the case is signed out while others prefer to archive digital files for much longer periods. In either case, the storage plan, budget, and architecture needs to be compatible with the clinical and administrative expectations. There may also be considerations for optimizing the location of the slide preparation and scanning facilities. Topics such as physical plant, environmental systems, network capacity, and redundant storage systems are all relevant to implementation.Understanding all of the upstream and downstream impacts of digital pathology is critical for successful implementation. Digitizing glass slides is the obvious change. However, more subtle changes include addressing upstream workflow changes in histology and downstream changes in transcription and in the LIS. Centralization of glass slides for scanning allows cases to be distributed electronically. However, the glass may still need to be readily available so that pathologists may request re-scans or physical glass slides if necessary. The LIS and PACS systems can be leveraged to provide insights into the daily workflows to anticipate busy periods and slide volumes. The LIS may need to accommodate and track cases signed out digitally for compliance and regulatory purposes.While the hardware, software, and infrastructure needs are critical, the people and change management issues are equally important, if not more so. Successful implementation of digital pathology requires a team approach and the support of many different stakeholders. A multidisciplinary team encompassing representatives from groups impacted by digital pathology would be well positioned to manage concerns and set expectations.

 

Learning Objectives

  1. Describe relevant scanner considerations for implementing digital pathology in various settings.         
  2. Discuss potential workflow implications for digital sign out.        
  3. Understand infrastructure requirements associated with adopting high volume scanning.
Chat bot