Digital Pathology 101 Chapter 4 | Digital Pathology Applications

Get the PDF of “Digital Pathology 101” Book here

Get the paper copy  of “Digital Pathology 101” on AMAZON

Have you started your digital pathology journey already?

Chances are that if you are reading this, you have. You have started it in a particular point of “digital pathology entry”. Maybe it was tissue image analysis, virtual rounds on whole slide images or validation of a scanner.

My “digital pathology entry point” was tissue image analysis and only through the lens of this application have I learned what are the other digital pathology applications.

In this chapter you will learn about all the current applications of digital pathology.

Because of where I started my journey I will always be biased towards tissue image analysis and AI, but revisiting the overview provided in this chapter will help me have all the other applications in mind, when I continue my journey of promoting digital pathology in the scientific and medical community.

I hope it will be a good basis for you as well. So let’s dive into the contents.

Here is what you will learn in Chapter 4 of the “Digital Pathology 101” book:

We’ll start by looking at the clinical applications. This includes

• the use of digital pathology for primary diagnosis in surgical pathology and cytopathology.

It facilitates more detailed examination and collaboration between pathologists. We’ll also discuss

• how telepathology enables remote intraoperative consultations and second opinion consults.

And we’ll touch on the

• education and training benefits, from resident teaching to continuing medical education.

Moving to research, we outline key applications like

• quantitative image analysis,
• AI and machine learning for predictive modeling, and
• high throughput analysis.
•  collaboration, allowing researchers to simultaneously access images
•  large scale studies and validation across institutions.

In drug development, digital pathology enhances

• preclinical histopathology and
• biomarker evaluation
•  clinical trials by eliminating slide shipment and enabling centralized review.

Digital tools can also assist in developing companion diagnostics, although regulatory requirements here are still evolving.

While each application has its challenges, the overarching benefit of digital pathology is its

• capacity to connect workflows,
• enhance efficiency, and
• open new possibilities across clinical, research, and drug development spheres.

Understanding the breadth of these applications provides a compass for navigating our own digital pathology journeys.

Enjoy this chapter and I’ll talk to you in chapter 5.

Support the show

Get the PDF of “Digital Pathology 101” Book here

Get the paper copy  of “Digital Pathology 101” on AMAZON

Become a Digital Pathology Trailblazer and See you inside the club: Digital Pathology Club Membership

• Digital Pathology 101 Book (PDF)
• Digital Pathology 101 Book on Amazon
• Digital Pathology Club Membership

CHAPTER 4: DIGITAL PATHOLOGY APPLICATIONS

 I. Introduction

Digital Pathology, by leveraging modern imaging technology and computational power, has brought transformative changes to the landscape of pathology practice, research, and drug development. This chapter will delve deeper into the wide array of applications that this technology has to offer. We will explore clinical applications, including primary diagnosis across various specialties, telepathology, and its role in education and quality assurance. Our attention will then shift to research applications, where digital pathology shines in digital image analysis, the use of artificial intelligence, machine learning, and high-throughput data management. The ability of digital pathology to foster collaboration and data sharing within the research community will also be highlighted. Next, we will investigate how digital pathology is catalyzing drug development. From preclinical studies to clinical trials, digital pathology aids in ensuring safety, efficacy, and the development of companion diagnostics. Alongside, we will shed light on the relevant regulatory aspects. We will conclude by discussing the current challenges and the future prospects of this vibrant field. Undeniably, digital pathology is making significant strides in advancing healthcare and this chapter aims to underline its diverse applications in propelling us towards a more efficient, accurate, and integrated future in pathology.

 II. Clinical Applications of Digital Pathology

A. Primary Diagnosis: Surgical Pathology and Cytopathology

One of the mainstays of digital pathology lies in its capacity for primary
diagnosis, notably in the realms of surgical pathology and cytopathology.

In surgical pathology, digitized slide images serve as the foundation for
comprehensive histopathological assessments. These images, taken from tissue
samples excised or biopsied during surgeries, are meticulously scrutinized by
pathologists for abnormalities. The utilization of digital pathology allows for
high-resolution, multi-plane examination, enhancing the accuracy of diagnoses, from the classification of cancer types to the grading of disease severity. Furthermore, it
facilitates collaborative reviews and consultations, thereby expediting the diagnosis
process.

Digital pathology also finds extensive application in cytopathology, which deals
with the examination of cells to detect diseases. Digitization of cytology specimens
such as Pap smears and fine-needle aspiration samples enables pathologists to
navigate easily through different planes of the specimen, zoom in for detailed analysis,
and utilize image analysis algorithms to detect subtle changes indicative of disease. By
augmenting the traditional microscope-based diagnosis, digital cytopathology enhances
diagnostic confidence and efficiency, contributing substantially to improved patient
care.

B. Telepathology: Intraoperative Consultations and Second Opinion
Consultations

In the realm of telepathology, digital technology is playing a pivotal role in intraoperative consultations and providing second opinions.

Intraoperative consultations, also known as frozen section diagnoses, are a critical part of many surgical procedures, where real-time pathological analysis is needed to guide surgical decisions. Through digital pathology, high-resolution images of the tissue under examination can be swiftly transferred to a remote pathologist for instant analysis. This not only facilitates immediate feedback to the surgical team, but it also allows access to subspecialty expertise, irrespective of geographic boundaries.

Moreover, digital pathology has revolutionized the process of seeking second opinions. Traditional practices of physically sending glass slides for review can be time-consuming and fraught with the risk of damage or loss. However, with digital pathology, images can be securely and instantaneously shared with expert pathologists anywhere in the world. This accessibility can prove invaluable, particularly for complex or rare cases where additional expert insight can significantly enhance diagnostic accuracy, leading to more effective treatment plans and better patient outcomes.

 C. Education and Training: Resident and Student Training and Continuing Medical Education

Digital pathology has emerged as an influential tool in medical education and training, particularly for pathology residents and students. Through digital platforms, students and residents can access a rich archive of high-resolution slide images that encompass a wide variety of diseases, thereby supplementing their learning with extensive real-world examples. Moreover, these digital libraries offer greater flexibility, allowing learners to study at their own pace and revisit challenging cases whenever they wish. This technology also makes it possible to annotate slides and share them for collaborative study and discussion, further enhancing the learning experience.

 Continuing Medical Education (CME) is another arena where digital pathology shines. As the field of pathology continually evolves with new diagnostic techniques, disease markers, and therapeutic implications, it is vital for practicing pathologists to keep up-to-date. Digital pathology platforms can host webinars, seminars, and interactive case studies, making it easier for pathologists to access and engage with the latest information and research. In particular, image analysis software, including artificial intelligence and machine learning algorithms, can be used in these educational settings to illustrate the latest advancements in pathology. In this way, digital pathology serves as an indispensable resource for lifelong learning, ensuring pathologists are equipped with the most current knowledge and skills in their field.

 D. Quality Assurance and Accreditation

Digital pathology plays a crucial role in enhancing the quality assurance (QA) process in clinical laboratories. It facilitates more efficient and standardized review of cases, supporting better reproducibility and consistency in pathology reporting. With digital pathology, slides can be easily archived and retrieved, making it convenient to revisit cases for quality control checks or additional consultations. These features allow for a robust QA system that is less prone to errors compared to conventional methods.

Furthermore, digital pathology can significantly assist in the accreditation process for pathology laboratories. Accreditation bodies often require documentation of QA processes, and the digital nature of these systems allows for streamlined and efficient record-keeping. For instance, tracking of slide review processes, analytical precision of digital tools, or the proficiency of pathologists in using digital pathology systems can be automated and easily reported.

Moreover, digital pathology supports inter-laboratory quality control measures, allowing different labs to compare results and ensure consistency. It also enables external quality assessment schemes, where digital slide sets can be shared widely with participating labs for standardized testing.

In summary, the digitalization of pathology boosts the effectiveness of quality assurance measures and simplifies the accreditation process. Therefore, it is instrumental in elevating the overall quality of pathology services.

E. Integration with Laboratory Information Systems (LIS) and Electronic
Health Records (EHRs)

Digital pathology is progressively intertwining with other critical healthcare IT systems, such as Laboratory Information Systems (LIS) and Electronic Health Records (EHRs), creating an interconnected environment for improved patient care. This integration facilitates the seamless flow of data and information across different platforms, thereby enhancing efficiency, reducing errors, and supporting comprehensive patient management.

When digital pathology systems are integrated with LIS, they provide an efficient conduit for exchanging diagnostic data, such as patient demographics, test requests, and results. This allows pathologists to access the necessary patient data directly within the digital pathology interface, fostering better informed and quicker diagnoses. Additionally, pathologists can directly upload their reports to the LIS, reducing transcription errors and improving turnaround times.

The integration of digital pathology with EHRs offers even more extensive
benefits. EHRs typically contain a patient’s entire health history, including previous
diagnoses, treatments, and medications. By linking digital pathology systems to EHRs,
pathologists gain immediate access to this crucial patient information, which can
greatly aid in diagnosis and treatment planning. Moreover, digital pathology results can
be swiftly transferred to the EHR, where they are immediately available for clinicians to
make timely treatment decisions.

In conclusion, the integration of digital pathology with LIS and EHRs enhances
the holistic view of a patient’s health, allows for more informed decisions, and ultimately leads to more efficient and effective patient care.

III. Research Applications of Digital Pathology

A. Digital Image Analysis: Quantitative Image Analysis and Biomarker
Research

In the realm of research, digital pathology’s application extends into profound areas of digital image analysis, including quantitative image analysis and biomarker research. These facets offer invaluable tools for gaining new insights into disease mechanisms and potential therapeutic targets.

Quantitative image analysis harnesses the power of digital pathology by providing a more objective, accurate, and reproducible method of analyzing tissue samples. With advanced algorithms, digital pathology software can quantify various morphological characteristics of tissue, such as the number, size, and shape of cells, their spatial relationships, and staining intensity. These data can then be used to draw conclusions about disease processes, patient prognosis, or treatment response. This not only mitigates the subjectivity inherent in manual assessments, but also enables analysis on a scale and with a level of detail that would not be feasible by human interpretation alone.

Digital pathology combined with brightfield and fluorescent multiplexing technology also plays a significant role in biomarker research. By using digital image analysis, researchers can identify and quantify biomarker expression in tissue samples with high precision. This can lead to the discovery of new disease markers and therapeutic targets. In conjunction with genomic data, digital pathology can also contribute to the development of personalized medicine by correlating morphological features with genetic alterations. Ultimately, the application of digital pathology in biomarker research can expedite drug discovery and enhance our understanding of disease biology.

 B. Artificial Intelligence and Machine Learning: Predictive and Prognostic Analysis

Artificial Intelligence and Machine Learning have emerged as game-changers in digital pathology, specifically in predictive and prognostic analysis. These advanced technologies have the potential to make the field of pathology even more precise, comprehensive, and efficient.

Predictive analysis in digital pathology involves the use of AI algorithms that can ‘learn’ from large datasets of labeled images to make predictions about new, unseen images. This might involve predicting the presence or absence of disease, the subtype of a tumor, or the likelihood of disease progression or response to a particular treatment. These models can analyze patterns and structures in histopathological images that might be too complex or subtle for the human eye to detect. Consequently, they can support pathologists in making more accurate and confident diagnoses and therapeutic decisions.

In prognostic analysis, AI and ML algorithms can be trained to predict patient outcomes, such as survival time or the likelihood of disease recurrence, based on features in the histopathology images. These prognostic models can integrate a variety of data types, including histopathological, clinical, and molecular data, to create a comprehensive view of the patient’s disease. By identifying patterns and correlations that might not be readily apparent to human observers, these models can provide additional insights to guide treatment decisions and inform patient counseling.

In summary, the application of AI and ML in digital pathology is not only enhancing our capacity to interpret complex histopathological images but also paving the way for more personalized and effective healthcare.

C. Data Management and High Throughput Analysis

The advent of digital pathology has resulted in the generation of large amounts of digital image data that require effective management and analysis. We are far beyond the point of putting all the image files in folders manually. This would lead to an organizational disaster.

 Data management in digital pathology includes storing, retrieving, and organizing digital pathology images and associated metadata. This necessitates the use of robust data management systems capable of handling large and complex data sets, ensuring data integrity, and providing user-friendly interfaces for data access and navigation. In plain words – it’s no longer enough to get a scanner and start digitizing. The image and data management system is a must and it’s often separate from your viewer and your image analysis software.

High throughput analysis is another major application of digital pathology in research. This involves the automated scanning and image analysis of large numbers of histopathological slides, enabling researchers to analyze many samples simultaneously. This can greatly increase the speed and efficiency of research studies, particularly in large-scale studies or when dealing with complex diseases that require the analysis of multiple tissue samples and multiple markers.

Moreover, high throughput analysis can also facilitate the discovery of new patterns and insights that might be missed in smaller or manually analyzed datasets. The ability to rapidly analyze large volumes of data can help identify trends, correlations, and outliers, providing a more comprehensive and accurate picture of the disease being studied (yes, I am thinking of cancer and immuno-oncology). In essence, effective data management and high throughput analysis are critical for maximizing the utility and impact of digital pathology in research.

D. Collaboration and Data Sharing in Research Settings

In the field of digital pathology, collaboration and data sharing are instrumental in broadening the scope and depth of research. Remember, collaboration was the reason telepathology (the “mother” of digital pathology) was developed in the first place. The digitization of pathology slides facilitates collaboration among researchers, pathologists, computer scientists and clinicians from different geographical locations and various institutions. They can remotely access and review the same sets of digital pathology images simultaneously, fostering more integrated and interactive discussions about disease processes, research findings, complex case reviews and conclusions drawn from tissue image analysis results.

Data sharing also allows researchers to validate their findings by using datasets from other studies or institutions, enhancing the reliability and generalizability of their results. In addition, it promotes a more open science approach by making the data accessible to the wider scientific community, encouraging the reuse of data, and stimulating new research ideas.

Furthermore, shared digital pathology databases or biobanks can serve as valuable resources for training artificial intelligence models, which require large amounts of data for training and validation. By facilitating data sharing, digital pathology is catalyzing advances in research, education, and clinical practice, paving the way for more collaborative, data-driven, and transparent research environments. 

IV. Drug Development Applications of Digital Pathology

A. Preclinical Studies

Digital pathology has found its significant place in the arena of preclinical studies, fundamentally transforming the process of drug development. For animal model and toxicologic safety studies, digital pathology expedites the process of histopathological analysis, allowing for quicker assessment of the effects of drugs on animal tissues. This provides valuable insights into the safety and efficacy of new drug candidates, which are critical for deciding whether a drug should advance to the next stage of development.

This capability is especially important as most pharmaceutical companies and many Contract Research Organizations (CROs) conducting those studies are multisite organizations with pathologists and researchers tasked to work together located in different countries, different continents, and different time zones. 

In the case of tissue cross-reactivity studies, digital pathology allows researchers to efficiently evaluate the potential reactivity of a therapeutic antibody with various human tissues. High-resolution digital images of tissue sections can be examined closely to determine the presence and location of antigens and assess the level of binding of a therapeutic antibody to these antigens. The ability to digitally store, share, and review these images can contribute to more consistent interpretation and more reliable results. Thus, digital pathology enhances the speed, accuracy, and efficiency of preclinical studies, thereby accelerating the drug development process.

 B. Clinical Trials

Traditionally, in multisite trials, the transportation of physical pathology slides from various sites to a central lab for review can be both time-consuming and risk-prone. This logistical challenge is effectively addressed by digital pathology (again, this was the reason Dr. Weinstein developed telepathology!).

Digital pathology allows for making pathology slides readily accessible for review from anywhere, without the need for physical transportation. This not only saves significant time but also reduces the risk of slide damage or loss. Furthermore, it allows for simultaneous access to slides by multiple pathologists, even if they are located in different geographical locations. This feature is particularly useful in multisite clinical trials, where several pathologists might need to review the same slides.  

In addition, digital pathology can enhance the consistency of pathological evaluations across different trial sites. This is because digital image analysis tools can be used to standardize the measurement and quantification of histological features, thereby reducing inter-observer variability.

Thus, by improving accessibility, reducing logistical issues, and enhancing consistency of evaluations, digital pathology can greatly benefit the logistics of multisite clinical trials.

Digital pathology also plays an indispensable role in biomarker-driven drug trials and the evaluation of drug safety and efficacy. In biomarker-driven trials, the use of digital pathology and image analysis enhances the ability to detect and quantify biomarkers in tissue samples. This is crucial for personalized medicine, as it allows for better stratification of patients and tailoring of treatments based on their biomarker profiles.

When it comes to safety and efficacy evaluation, digital pathology offers several advantages over traditional methods. It enables more efficient and accurate analysis of histopathological changes in tissue samples, which are often critical endpoints in clinical trials. High-resolution digital images allow for detailed examination and more precise measurements, which can contribute to more accurate assessments of drug effects. Additionally, the use of digital pathology facilitates the standardization and reproducibility of assessments, reducing the variability and potential bias in the results.

Furthermore, the digital format allows for easier storage, retrieval, and sharing of images, which can enhance the efficiency of the review process and enable collaboration among pathologists and other stakeholders. This can lead to better decision-making in clinical trials and ultimately contribute to the development of safer and more effective drugs.

 C. Companion Diagnostics Development

Digital pathology is instrumental in the development of companion diagnostics, which are laboratory or diagnostic tests used to determine the suitability of a particular drug for a specific patient. These diagnostics are essential for the concept of personalized medicine, where treatments are tailored to the individual patient based on their unique genetic and molecular profile.

Through the detailed analysis of digital pathology slides, researchers can uncover specific disease markers or patterns that correlate with a patient’s response to a particular drug. This information can then be used to develop diagnostic tests that identify these markers or patterns in future patients, effectively predicting their response to the drug.

Moreover, the integration of artificial intelligence and machine learning in digital pathology can automate and enhance this process. For example, AI algorithms can be trained to identify subtle histological features that might be overlooked by human observers, thereby improving the accuracy of the companion diagnostic.

By facilitating the development of companion diagnostics, digital pathology plays a critical role in advancing personalized medicine, where treatments are tailored to the unique molecular profile of each patient. This not only improves patient outcomes but also contributes to the overall efficiency of healthcare systems by ensuring that treatments are directed towards those who are most likely to benefit from them.

 D. Regulatory Aspects and Challenges in Drug Development

Drug development, especially the later development stages, happens in a highly regulated environment. Thus the utilization of digital pathology in drug development introduces a unique set of regulatory considerations and challenges. Regulatory bodies, such as the U.S. FDA and the EMA in Europe, require rigorous validation and standardization of the tools and processes used in drug development to ensure safety and efficacy. In the case of digital pathology, this includes validation of the digital pathology systems themselves, as well as the image analysis algorithms and artificial intelligence applications that may be used in conjunction with them.

The digitization of pathology also introduces new complexities with regards to data privacy and security. As patient information and sensitive data are digitized and shared for drug development purposes, robust measures must be in place to ensure the data’s security and compliance with privacy regulations like the HIPAA in the U.S. or GDPR in the EU.

Further, the incorporation of AI and machine learning tools into digital pathology introduces additional regulatory considerations. These technologies when trained end-to-end (from image input directly to result in contrast to performing the analysis stepwise) often function as ‘black boxes’, with their decision-making processes being largely opaque. Regulators are currently grappling with how to approach the validation and oversight of these tools, to ensure they are reliable, trustworthy, and do not inadvertently introduce bias or error into the drug development process.

The regulatory landscape for digital pathology in drug development is therefore complex and evolving. It requires ongoing collaboration between industry, regulatory bodies, and other stakeholders to establish appropriate standards and guidelines that can facilitate the potential benefits of digital pathology while managing these challenges.

That being said, regulatory compliance can be achieved and this will drive further adoption of this technology. We will talk more about this aspect in the next chapter on digital pathology applications in toxicologic pathology.

V. Conclusion

Digital pathology is a rapidly advancing field, poised at the forefront of transformative change in healthcare. Despite the considerable advancements and potentials, digital pathology is not without its challenges. These range from technical aspects such as data storage and computational needs, to more complex issues related to standardization, validation, and regulatory considerations. Equally, the integration of artificial intelligence and machine learning presents both unprecedented opportunities and new complexities. The ongoing development of these technologies, however, represents a promising future for overcoming current obstacles.

As we celebrate the potential of digital pathology, it is crucial not to underestimate the obstacles and the complexity of transitioning to digital operations. Currently, there is no universally applicable blueprint (and there might never be…) for implementing digital pathology, and the scientific literature is constantly evolving with new experiences and insights. This transformation will undoubtedly involve a lot of experimentation and problem-solving. It is imperative to remember that while the path to digital pathology is laden with potential obstacles, these hurdles are not insurmountable.

The importance of a dedicated, resilient, and creative digital implementation team cannot be overstated (you DO need a super enthusiastic, competent team that loves troubleshooting!) . The challenges that lie ahead will require team members who can rise to the occasion, persevere through challenges, and adapt to a rapidly evolving landscape. In the long run, overcoming these hurdles will position digital pathology as an integral part of future healthcare, enhancing diagnostic accuracy, research capabilities, and drug development processes. It is a journey that demands determination, but the rewards are promising and pivotal in advancing healthcare.

The role of digital pathology in advancing healthcare is multifaceted and significant. In the clinical realm, it stands to improve diagnostic accuracy, efficiency, and reach, particularly in underserved areas, through telepathology. In education, it provides a dynamic and versatile tool for training the next generation of pathologists. The research applications of digital pathology are profound, providing powerful tools for quantitative analysis, high throughput studies, and the leveraging of AI to uncover new insights. The application of digital pathology in drug development offers the potential to streamline the process and increase the precision of preclinical and clinical studies. 

As digital pathology continues to evolve, it will undoubtedly play an increasingly central role in shaping the future of healthcare, opening up new possibilities for patient care, medical research, and the development of therapeutic interventions. This promising field stands not only as a testament to the power of technological innovation but also to the potential for that innovation to substantially enhance our ability to understand, diagnose, and treat disease.

• Reimbursement for digital pathology in the clinic – how does that work? w/ Esther Abels, Visiopharm
• Lung PD-L1 AI: Development and Validation of an Automated PD-L1 Tumor Proportion Scoring Algorithm in Non-Small Cell Lung Cancer
• Why and how should pathologists keep up with AI? w/ David Harrison, University of St. Andrews
• Digital Pathology 101 Chapter 1 (Part 2) | Are Pathologists at Risk in the Digital Age?
• Podcast
• Digital Pathology 101 Chapter 3 | Image Analysis, Artificial Intelligence, and Machined Learning in Pathology