The first digital pathology platform receiving FDA approval has been introduced to the market and is leading to expanded deployment of hardware solutions, along with accompanying software tools and applications. These include slide-free imaging, enterprise-wide solutions and algorithms that use machine learning and AI to improve diagnoses, guide treatment, and address medical challenges across a variety of diseases. The merging of different fields including radiology and pathology is leading to ways of creating digital platforms and streamlining workflows. Fears stemming from how the introduction of these computational tools will impact the pathologist will be addressed, along with a review of emerging companies and approaches seeking to create the pathology diagnosis of the future.

Final Agenda

Scientific Advisory Board

Eric F. Glassy, MD, FCAP, Medical Director, Affiliated Pathologists Medical Group

Richard M. Levenson, MD, Professor and Vice Chair for Strategic Technologies, Department of Pathology & Laboratory Medicine, UC Davis Medical Center

Liron Pantanowitz, MD, Professor, Pathology & Biomedical Informatics, University of Pittsburgh Medical Center, Conference Chairman

David L. Rimm, MD, PhD, Professor, Pathology, Yale University School of Medicine

Monday, March 11

10:30 am Conference Program Registration Open (South Lobby)

OPENING KEYNOTE SESSION
10

11:50 Chairperson’s Opening Remarks

Eric F. Glassy, MD, FCAP, Medical Director, Affiliated Pathologists Medical Group

12:00 pm Advancing Cancer Diagnostics with Artificial Intelligence

Martin Stumpe, PhD, Senior Vice President, Data Science, Tempus

Rendering cancer diagnoses from tissues is a highly complex process that requires many years of expert training. It involves challenging tasks for pathologists, such as identifying rare events in very large images (e.g., micrometastases in lymph nodes), or classifying subtle differences between normal vs. tumor, or amongst similar-looking tumors that have very different treatment plans. These tasks are typically more challenging for human cognition, and, consequently, over- and under-diagnoses are not uncommon, resulting in non-optimal treatment selection. In recent year, deep learning has been successfully applied for several image-based tasks in digital pathology. This talk will discuss the potential of machine learning in cancer diagnostics, as well as the challenges to bring it to practical, and how those challenges can be overcome.

12:30 QuPath: Digital Pathology for Everyone

Peter Bankhead, PhD, Senior Lecturer, Digital Pathology, University of Edinburgh

QuPath is an open source digital pathology platform that is rapidly becoming established as the software of choice for many researchers worldwide working with whole slide images, both in academia and industry. This talk describes the background to QuPath, its current applications and latest developments, and discusses what the software offers pathologists, biologists and algorithm developers looking to apply sophisticated and reproducible analysis and visualization methods to complex tissue images.

1:00 Enjoy Lunch on Your Own

SLIDE-FREE IMAGING
10

2:30 Chairperson’s Remarks

Richard M. Levenson, MD, Professor and Vice Chair for Strategic Technologies, Department of Pathology & Laboratory Medicine, UC Davis Medical Center

2:40 Rapid, Automated Brain Tumor Diagnosis Using Stimulated Raman Histology

Daniel A. Orringer, MD, Assistant Professor, Neurosurgery; Member, Cancer Center, University of Michigan

Accurate, diagnostic histologic information is essential for decision making during brain tumor surgery. Conventional histologic techniques complicate and delay the access of clinicians to microscopic data. Stimulated Raman histology (SRH), a label-free method for generating histologic images of fresh specimens puts microscopic data at the surgeons’ fingertips. Moreover, SRH images are natively digital and well-suited to automated interpretation through emerging artificial intelligence (AI) methods. SRH and AI can be combined to unlock and annotate histologic information that can be leveraged by surgeons to provide the best possible care to their patients.

3:10 3D Light-Sheet Microscopy for Next Generation Pathology

Nicholas P. Reder, MD, MPH, Genitourinary Pathology Fellow, CAP In-Vivo Microscopy Committee, Junior Member, Harborview Medical Center, Department of Pathology, University of Washington

Light-sheet microscopy, in combination with optical clearing techniques, enables non-destructive 3D imaging of tissue. These 3D microscopic images yield new biologic insights into prostate carcinoma, kidney vasculature, and other 3D structural features. In addition, the non-destructive nature of this technique creates laboratory workflow efficiencies and better integration with genetic sequencing assays. I will give an overview of the current state of the technology, future directions, and potential for clinical adoption.

3:40 Deep Learning-Enabled Computational Imaging in Pathology

Aydogan Ozcan, PhD, Chancellor’s Professor, Electrical & Computer Engineering Bioengineering, UCLA; HHMI Professor, Howard Hughes Medical Institute; Associate Director, California NanoSystems Institute (CNSI); Founder of Holomic/Cellmic LLC

Deep learning is a class of machine learning techniques that uses multi-layered artificial neural networks for automated analysis of signals or data. Beyond its main stream applications such as the recognition and labeling of specific features in images, deep learning holds numerous opportunities for revolutionizing image formation, reconstruction and sensing fields. In this presentation, I will provide an overview of some of our recent work on the use of deep neural networks in advancing computational microscopy systems, also covering their applications in pathology.

4:10 Big Picture - Deep Analysis. Cloud Based Active Deep Learning for Computational Pathology

Thomas Westerling-Bui, PhD, Senior Scientist, Aiforia Inc.

Traditionally the use of DeepLearning in image analysis has been the domain of the computer and data scientists. Here we show how you can by simple deployment of Aiforia Cloud Solutions, enable yourself to train, validate, and deploy DeepLearning algorithms in your own digital pathology workflow. 

4:40 Refreshment Break and Transition to Plenary Session

5:00 Plenary Keynote Session (Room Location: 3 & 7)

6:00 Grand Opening Reception in the Exhibit Hall with Poster Viewing

7:30 Close of Day

Tuesday, March 12

7:30 am Registration Open and Morning Coffee (South Lobby)

8:00 Plenary Keynote Session (Room Location: 3 & 7)

9:15 Refreshment Break in the Exhibit Hall with Poster Viewing

PREDICTIVE IMMUNO-ONCOLOGY BIOMARKERS AND DIGITAL PATHOLOGY
10

10:15 Chairperson’s Remarks

David L. Rimm, MD, PhD, Professor of Pathology and of Medicine (Medical Oncology); Director of Pathology Tissue Services; Director of Translational Pathology, Yale University School of Medicine

10:25 Companion Diagnostics for the Adjuvant Setting for Immune Therapy

David L. Rimm, MD, PhD, Professor of Pathology and of Medicine (Medical Oncology); Director of Pathology Tissue Services; Director of Translational Pathology, Yale University School of Medicine

Companion diagnostics for immune checkpoint blockade (ICB) therapy have all been designed and executed by the pharma companies producing the drugs. However, in the adjuvant setting, better markers are needed. In the adjuvant setting, since the side effects are toxic and only 1 in 5 benefits, biomarkers that predict which patients do NOT need drug are most important. Here we will discuss biomarkers with promise for selection in the adjuvant setting for ICB therapy.

10:55 Harnessing Digital Pathology Tools to Select Early Stage Melanoma Patients for Adjuvant Immunotherapy

Yvonne Saenger, MD, Department of Medicine, Division of Hematology/Oncology; Director, Melanoma Immunotherapy, Columbia University

Better biomarkers are needed in early stage disease because the toxicities of immunotherapy are of greater concern in healthy patients. While it is known that tumor infiltrating lymphocytes (TILs) confer a favorable prognosis, the subjectivity and lack of precise quantification of this metric has limited application to clinical practice. Quantitative multi-plexed immune fluorescence (QmIF) provides a means to both phenotype and reproducibly quantify TILs while also analyzing their spatial position within the tumor. We present data showing that the ratio of CD8+CD3+ T cells to CD68+ macrophages within the stroma of primary melanoma tumors correlates with clinical outcome and propose that this method has applicability to a variety of cancer types for biomarker development.

11:25 From Tissue-Exhausting to Tissue-Sparing Companion Diagnostics through Machine Learning

Kimary Kulig, PhD, MPH, Head of Oncology Research, Verily Life Sciences

Peter Cimermancic, PhD, Scientist, Computational Biology, Verily Life Sciences

Clinical presentation of non-small cell lung cancer (NSCLC) is often metastatic/late-stage. Surgical resection or core needle biopsy material is absent/limited in ~40% of cases. The current tissue requirement for diagnosis and treatment-guiding biomarker testing is approaching 20 slides. With many therapies in development for lung cancer, the current biomarker testing paradigm is unsustainable. Our work explores how companion diagnostics may be replaced by image processing tools and machine learning.

11:55 Artificial Intelligence for Immuno-Oncology Pathology:
From Discovery to AI-Powered Companion Diagnostics

Andrew Beck, MD, PhD, CEO, PathAI

Pathology plays a central role in the field of immuno-oncology (IO). Recent advances in artificial intelligence and computer vision offer tremendous potential for discovering new pathologic mechanisms of IO treatment response, identifying new signatures for patient selection, and enabling the direct integration of clinical, transcriptomic and pathology data. We will discuss these new advances and their role in accelerating progress in the development and approval of IO therapies for cancer patients.

12:25 pm Enjoy Lunch on Your Own

1:35 Refreshment Break in the Exhibit Hall with Poster Viewing

STATE-OF-THE-ART CLINICAL PRACTICE USING DIGITAL PATHOLOGY
10

2:05 Chairperson’s Remarks

Liron Pantanowitz, MD, Professor, Pathology & Biomedical Informatics, University of Pittsburgh Medical Center

2:10 NSH/DPA Online Digital Pathology Certificate of Completion

Elizabeth A. Chlipala, BS, HTL(ASCP)QIHC, Laboratory Manager, Premier Laboratory, LLC

As digital pathology matures, it is apparent that we need well-trained individuals to manage our whole slide imaging systems. This discussion will introduce the joint NSH/DPA online self-paced Digital Pathology Certificate Program which was released in May 2018. An overview of how this program was developed, the content of the educational modules, and the manner in which this program is being provided will be presented. In addition, there will be discussion on the importance of this educational program to the field of digital pathology.

2:40 Integrated Diagnostic Reporting: Improving the Diagnostic Product by Real-Time Integration of Radiology and Pathology Studies

W. Dean Wallace, MD, Professor, Pathology and Laboratory Medicine, David Geffen School of Medicine, University of California, Los Angeles

The increase in complexity of new testing modalities in pathology and radiology has led to new challenges in gathering all essential information and forming a coherent diagnostic message for managing clinicians. Opportunities for optimized patient care may be lost and diagnostic errors may occur without a rigorous assimilation of the various testing modalities in pathology and radiology. This presentation will discuss the challenges of pathology and radiology correlation in the current diagnostic paradigm and the numerous opportunities for improved workflows and patient outcomes with an effective diagnostic integration process. This presentation will also review the solution created by researchers and doctors at UCLA that automates the diagnostic integration process and the clinical practice experience.

3:10 Value-Added Digital Pathology and Incremental Implementation

Stephen M. Hewitt, MD, PhD, CAPT, USPHS, Head, Experimental Pathology Laboratory, Laboratory of Pathology, Center for Cancer Research, National Cancer Institute, National Institutes of Health

Adoption of Digital Pathology is dependent on a value-added model. The conversion of a manual to digital workflow alone is insufficient to overcome the operational cost. Value-added paradigms for digital pathology include Image Analysis, Computer-Aided Diagnosis, and add-on utilities to improve performance of downstream assays such as Molecular Pathology. Ultimately, adoption of Digital Pathology requires an incremental approach with the pathology laboratory based on value-driven proposition.

3:40 2021: A Path Odyssey — One Lab’s Experience in Digital Transformation

Matthew O. Leavitt, MD, CMO & Founder, LUMEA

Seven years ago, our community-based pathology lab embarked on a journey to implement an end-to-end digital workflow. We learned that to reap the real value of digital transformation, a lab must rethink the entire diagnostic workflow, retool parts of the process amenable to automation, and redeploy valuable lab personnel to fill new roles. LUMEA’s mission is to encourage adoption of digital pathology by providing labs with tools that unleash digital pathology’s potential to improve diagnostic quality, while lowering cost. Our clinical experience implementing the following digital pathology tools will be discussed:

4:10 St. Patrick’s Day Celebration in the Exhibit Hall with Poster Viewing

5:00 Breakout Discussions in the Exhibit Hall

Utilizing Whole Slide Imaging and Image Analysis in the Histology Laboratory for Quality Assurance and Improvement

Elizabeth A. Chlipala, BS, HTL(ASCP)QIHC, Laboratory Manager, Premier Laboratory, LLC

• The importance of standardization and quality process improvement in histotechnology and its significance to the success of implementing a digital pathology solution
• Utilization of digital pathology technology to improve overall efficiency and quality of the histology preparations
• Utilization of digital pathology technology to document the accuracy and precision of a histology process

Digital Pathology Interoperability Priorities

Raj C. Dash, MD, Professor and Vice Chair, Pathology IT, Duke University Health System; Medical Director, Laboratory Information Systems

• Describe the needs of your organization as a consumer of or contributor to a larger digital pathology environment
• Discuss the current level of interoperability among current market offerings in Digital Pathology
• Help prioritize the efforts of standard setting and interoperability initiatives for Digital Pathology

Digital Pathology Applications for Predictive Biomarkers

Ehab A. ElGabry, MD, Senior Director, Pathology & Companion Diagnostics, Pharma Services Medical Director, Ventana Medical Systems, Inc., A Member of the Roche Group

• Multiplexing
• Digital training and electronic learning modules
• Clinical scoring algorithms development

Practical Considerations when Implementing Digital Pathology

Douglas J. Hartman, MD, Associate Professor of Pathology and Director, Division of Pathology Informatics, University of Pittsburgh Medical Center, Ventana Medical Systems, Inc., A Member of the Roche Group

• Presenting digital pathology to different stakeholders in your institution
• Establishing a return on investment for digital pathology
• Unique capabilities that can only be offered thru digital pathology

Pathology AI: The Promise and the Problems

Michael C. Montalto, PhD, Vice President, Pathology and Clinical Biomarker Laboratories, Translational Medicine, Bristol-Myers Squibb

• Application of AI to pathology in translational medicine
• Pathology AI based companion and complementary diagnostics
• Barriers to adoption of pathology AI in clinical practice

AI and ML in Pathology Practice

Hooman H. Rashidi, MD, FASCP, Professor and Vice Chair, GME, Director of Residency Program; Director, Flow Cytometry & Immunology, Department of Pathology and Laboratory Medicine, University of California, Davis School of Medicine

• Data types
• Types of ML platforms
• Validation of ML platforms in pathology

6:00 Close of Day

Wednesday, March 13

7:30 am Registration Open and Morning Coffee (South Lobby)

8:00 Plenary Keynote Session (Room Location: 3 & 7)

10:00 Refreshment Break and Poster Competition Winner Announced in the Exhibit Hall

NOVEL APPLICATIONS OF DIGITAL PATHOLOGY
10

10:50 Chairperson’s Remarks

Liron Pantanowitz, MD, Professor, Pathology & Biomedical Informatics, University of Pittsburgh Medical Center

11:00 The Application of Pathology-Based Image Analysis and AI to Patient Selection Strategies in Immuno-Oncology Clinical Trials

Michael C. Montalto, PhD, Vice President, Pathology and Clinical Biomarker Laboratories, Translational Medicine, Bristol-Myers Squibb

Understanding the role of the tumor micro environment is paramount to the success of patient selection strategies in immune-oncology drug development. This talk will explore the role of automated image analysis and artificial intelligence for pathology samples in I-O drug development, and its potential as a platform for patient stratification and companion diagnostics.

11:30 Building the Digital Pathology Cockpit: What Is It Looking Like and Where Does It Need to Go?

Douglas J. Hartman, MD, Associate Professor of Pathology and Director, Division of Pathology Informatics, University of Pittsburgh Medical Center

Digital pathology is increasing in adoption throughout pathology departments. There is a transition of the working environment from a microscope environment to a computer. This creates the opportunity to augment the tools that are available for a pathologist in order to support better diagnostics. These additional tools have been described as a diagnostic cockpit. There are lessons that can be learned from our radiology colleagues as the pathology cockpit evolves.

12:00 pm Digital Pathology Applications for Personalized Health Care: An Industrial Perspective on Immunohistochemical Assays

Ehab A. ElGabry, MD, Senior Director, Pathology & Companion Diagnostics, Pharma Services Medical Director, Ventana Medical Systems, Inc., A Member of the Roche Group

In an era focused on personalized health care (PHC), the use of immunohistochemical assays for targeted immunotherapies (e.g. Programmed death ligand 1) is rapidly expanding. The objective of this presentation is to give a brief overview of the current and future role of digital pathology in the development process of these assays from early hypothesis to launch. More specifically we will share our experience using image analysis software, digital platforms for training, artificial intelligence and multiplexing technologies.

12:30 Enjoy Lunch on Your Own

1:10 Refreshment Break in the Exhibit Hall and Last Chance for Poster Viewing

BUSINESS-RELATED ASPECTS OF DIGITAL PATHOLOGY
10

1:50 Chairperson’s Remarks

Liron Pantanowitz, MD, Professor, Pathology & Biomedical Informatics, University of Pittsburgh Medical Center

2:00 The Impact of FDA Approval on Whole Slide Imaging

Liron Pantanowitz, MD, Professor, Pathology & Biomedical Informatics, University of Pittsburgh Medical Center

Since the FDA’s approval of the first whole slide imaging (WSI) system for primary diagnosis in surgical pathology, there has been considerable commotion in the digital pathology community. This talk will review the impact of the FDA allowing Philips to market their WSI system for digital pathology. The implications for pathology laboratories, vendors, and the digital pathology market will be discussed. Future regulatory challenges related to adding new components and machine learning tools to FDA-cleared WSI systems will also be reviewed.

2:30 Beyond Technical Performance: Some Additional Considerations That May Shape the Future of AI in Medicine

Richard M. Levenson, MD, Professor and Vice Chair for Strategic Technologies, Department of Pathology & Laboratory Medicine, UC Davis Medical Center

There is much more to consider than just technical performance when contemplating the application and acceptance of new tools in mature medical fields. In the case of AI, there are issues to address across many domains. Legal concerns include regulatory aspects and the impact of malpractice fears; financial considerations suggest that if costs are to be contained, then care-generated revenue will have to be partitioned across animate and inanimate providers; and the evolution of a skilled workforce in an AI-infused world is presently hard to foresee.

3:00 Integrated Computational Pathology at the Memorial Sloan Kettering Cancer Center: An Overview.

Peter Schueffler, PhD, Machine Learning Scientist, Memorial Sloan Kettering Cancer Center

The Memorial Sloan Kettering Cancer Center is increasing the effort to digitize their pathology slides, currently scanning 40k slides per month. This enables the development of clinical grade machine learning models which profit from the large amount of data and MSKCC’s large computer cluster. We integrated a universal slide viewer for the clinical and research workflows, to facilitate the viewing and annotation of digital slides within different systems in the hospital. This talk gives an overview of this setup as well as the current research projects in computational pathology.

3:30 Session Break

DIAGNOSTICS OF THE FUTURE
10

3:40 Chairperson’s Remarks

Eric F. Glassy, MD, FCAP, Medical Director, Affiliated Pathologists Medical Group

3:45 What is the Future of Pathology?

James M. Crawford, MD, PhD, Professor and Chair, Department of Pathology and Laboratory Medicine, Senior Vice President of Laboratory Services, Northwell Health

An article-of-faith for pathologists, and pathology in general, is that we practice high impact patient-centered healthcare that touches virtually every person in their life journey. But to policy-makers and our stakeholders, including consumers, pathology is a commodity. And in the transformation of American healthcare to a value-based system, pathology is at risk of being left off the agenda, except perhaps for cancer screening programs. While laboratory medicine and precision medicine are getting a good run at building their “value-added” evidence base for the future practice of medicine, anatomic pathology has not yet done so. Digital pathology and AI/ML hold promise, but if the only achievement is facilitating more effective workflow for pathologists, then we will have fallen short. This talk will examine how pathology (as in anatomic pathology) might effectively bring its “value statements” into the future of healthcare and secure its future position as a valued asset.

4:15 Interoperability Advances in Digital Pathology and Implications for Your Organization’s Imaging Strategy

Raj C. Dash, MD, Professor and Vice-Chair, Pathology IT, Duke University Health System; Medical Director, Laboratory Information Systems

The Pathology and Lab Medicine (PALM) domain of the Integrating the Healthcare Enterprise (IHE) organization has collaborated with DICOM (Digital Imaging and Communications in Medicine) working group 26 to propose an initial draft solution supporting an environment of interoperability among vended digital pathology instrumentation and systems focused on acquisition of digital assets critical for anatomic pathology diagnostics. This solution aims to increase the value of all vended digital pathology components.

4:45 Artificial Intelligence (AI)/Machine Learning (ML) Expert Systems in Pathology

Hooman H. Rashidi, MD, FASCP, Professor and Vice Chair, GME, Director of Residency Program; Director, Flow Cytometry & Immunology, Department of Pathology and Laboratory Medicine, University of California, Davis School of Medicine

AI/ML platforms will undoubtedly become an integral part of our future medical practice workflow (such as pathology and radiology). These ML platforms can help the workflow with triage, help formulate a differential diagnosis, help verify one’s diagnostic impression, and enhance the laboratory workflow as the number of cases per diagnostician increases. The development and validation of such platforms requires a multidisciplinary team that includes a specialist, machine learning expert(s) and certain digital delivery platform experts.

5:15 Close of Conference Program