Back to the Future of Digital Pathology: From Compatibility to Interoperability

Introduction: The Crossroads of Digital Pathology

Digital pathology is not just a technological upgrade; it is one of the deepest transformations taking place in modern medicine. By moving from the traditional microscope to an ecosystem of high-resolution images and artificial intelligence, we are changing how diagnosis is practiced, how data circulates, and ultimately how patients are treated.

Yet, at the heart of this evolution lies a subtle but decisive distinction, often misunderstood even among experts: the difference between compatibility and interoperability. This is not merely a technical nuance. It is a strategic decision that will determine whether a healthcare system can innovate, scale, and fully deliver on the promise of digital medicine.

To make the distinction clearer, think of everyday examples. Compatibility is like buying a solid, well-crafted door. If it fits the frame and opens and closes smoothly, it is compatible. Interoperability, on the other hand, is making sure that this door is also connected to the home’s alarm system, the smart lighting, and the digital lock. It is about ensuring systems not only work, but also work together within a larger ecosystem.

A similar analogy comes from the gaming world. A video game is compatible if it can run on Windows or Linux. But if the same game could share your progress and achievements with another game (even from a different company) that would be interoperability. It requires a common protocol, a shared language, and a willingness to go beyond “just working.” (Confession: I also keep a small collection of video games; if you ever want to talk about AI and gaming, my inbox is open.)

This distinction is particularly urgent in oncology. Cancer pathology demands speed and precision that traditional workflows can no longer guarantee. With fewer pathologists available and the pressure of precision medicine, digital pathology adoption is accelerating. The first wave of adoption focused on compatibility, mainly scanning slides for easier visualization, fewer broken glass slides, and the ability to review cases remotely. These are real benefits, but the reliance on proprietary file formats has created data silos that now block integration with other hospital systems (LIMS, EHRs) and limit the power of artificial intelligence.

The true value of digital pathology is not simply to view slides on a screen. It lies in unlocking data at scale, to drive advanced analytics, enable multicenter collaboration, and transform patient care. This is why the debate is not academic. It is a strategic crossroad: whether to settle for short-term compatibility or to invest in long-term interoperability.

Compatibility: A Partial Entry Point

In technology, compatibility is usually defined as the ability of a system, component, or device to function alongside another. In digital pathology, this means that a viewer can read a whole-slide image (WSI) generated by a scanner from a different vendor. For example, Aperio produces .svs files, Hamamatsu uses .ndpi, and Philips relies on its proprietary iSyntax format. Compatibility shows up when a software viewer can interpret these formats, allowing the pathologist to look at the images without being tied to the scanner’s original software.

This approach has clear, immediate benefits and was a major driver of early adoption. For hospitals and labs, compatibility enabled remote and collaborative access to slides pathologists could review cases from anywhere without needing to physically ship fragile glass slides. It also improved workflows: turnaround times dropped, and common operational errors (broken slides, mislabeling, lost cases) became less frequent. Many viewers also added simple but valuable tools: live zooming, annotations for team discussion, and area measurements. Together, these changes represented a leap in convenience and efficiency.

However, these short-term gains often hide structural limitations. The most critical issue is the loss of metadata, the “data about the data,” such as staining, magnification, scanning date, or patient identifiers. Proprietary formats tend to encapsulate this information in non-standard ways. When a “compatible” viewer opens them, key metadata can be lost or require manual entry. This not only increases the risk of clinical errors but also compromises data integrity.

What emerges is the creation of data silos. Patient history may sit in the electronic health record (EHR), lab results in the laboratory information system (LIS), and pathology slides in a proprietary image management system (IMS). Each system is functional on its own, but they don’t talk to each other. For the clinician, this fragmentation makes it nearly impossible to have a complete, unified view of the patient—sometimes leading to duplicate tests, incomplete diagnoses, or missed insights.

Another limitation is the lack of bidirectional integration. Compatibility may let you view a slide, but it does not ensure that your digital pathology system can automatically exchange information with the LIS or EHR. In practice, this forces lab staff to re-enter data manually across multiple systems a process that is time-consuming, prone to error, and administratively costly.

In short, compatibility opens the door to digital pathology, but it leaves you standing in the hallway. You can see the slides, but you are not yet inside the connected house of healthcare data.

Interoperability: The Real Integration

Interoperability goes beyond coexistence. It is the ability of systems, protocols, and technologies to let data flow securely, in real time, and without manual intervention across different platforms. Unlike compatibility (which is often one-directional and limited) interoperability means seamless, meaningful communication between systems, so that clinical information is not just stored but actively integrated into patient care.

It can be understood at different levels, each building on the previous one:

• Foundational interoperability: the basic ability to transfer data from one system to another (like sending a PDF by email). The receiving system gets the file but cannot “understand” its contents.

• Structural or syntactic interoperability: defines how data is organized, so the receiving system can identify fields such as name, date, or lab value. This is where standards like DICOM play a central role.

• Semantic interoperability: ensures that data carries the same meaning across systems. This requires shared vocabularies and coding standards such as ICD-10 for diagnoses or LOINC for lab results.

• Organizational interoperability: the highest level, where data can move across institutions with different workflows, regulations, and governance structures. This is what enables networks of hospitals to function as a coordinated ecosystem.

For digital pathology, the cornerstone is DICOM Supplement 145, which specifies how whole-slide images (WSI) should be encoded, stored, and shared. With this standard, pathology images can be integrated into the same PACS(Picture Archiving and Communication System) that radiology has used for decades. Other standards complement DICOM: HL7 connects non-image health data (lab records, patient history), and its modern evolution FHIR provides API-based integration for real-time applications, ideal for embedding AI algorithms. Meanwhile, IHE (Integrating the Healthcare Enterprise) defines practical “integration profiles” to ensure that systems using DICOM and HL7 work together coherently in clinical workflows.

The benefits of interoperability reach far beyond visualization:

• Clinically, it offers a holistic view of the patient. Pathologists and oncologists can access slides, lab values, and medical history in a single environment, reducing the risk of errors, avoiding duplicate tests, and improving care coordination.

• Operationally, it automates workflows, eliminates double data entry, and reduces administrative burden, freeing up time for clinicians and lab staff.

• Strategically, it lays the foundation for AI and big data: enabling multicenter research, faster algorithm deployment, and scalable infrastructures for personalized medicine.

We already have real-world evidence. At Stanford Medicine, for instance, a platform was deployed that converts proprietary formats into DICOM, making it possible to integrate pathology images into the hospital PACS, LIS, and AI tools. This interoperability streamlined clinical decision-making and accelerated adoption of AI-driven insights. Similarly, hospitals with bidirectional integration between LIS and image management systems have reported fewer workflow bottlenecks and greater consistency in case management.

In essence, interoperability is not about “making the system work.” It is about building a connected ecosystem where data flows, speaks the same language, and drives both clinical and organizational value.

The Risks of Staying at the Level of Compatibility

While compatibility offers an easy entry point into digital pathology, it carries significant long-term risks, what engineers often call “technical debt.” At first it feels convenient, but over time it locks institutions into expensive, rigid systems that limit growth, collaboration, and patient safety.

From an economic perspective, the biggest danger is vendor lock-in. Imagine investing in a scanner that only produces files readable by the same company’s viewer. Negotiating better prices or integrating solutions from other vendors becomes nearly impossible. Migrating later to an open, interoperable system means converting massive archives of proprietary data into DICOM, a process that is slow, expensive, and highly error-prone. Hidden operational costs also add up: when lab staff must re-enter data into multiple systems because the digital pathology platform doesn’t talk to the LIS or EHR, administrative overhead skyrockets.

From a clinical perspective, compatibility can compromise patient safety. Without standardized metadata, critical details such as staining protocols, magnification, or patient identifiers may be lost or misrecorded during transfers. In non-interoperable systems, there is no unique patient identifier across platforms, raising the risk of mislinked or missing data. Fragmented information trapped in silos can easily lead to incomplete clinical histories, repeated tests, or even diagnostic errors.

From a strategic perspective, compatibility is a ceiling that blocks innovation. Scaling a digital pathology network across multiple hospitals becomes unfeasible if each one uses different proprietary formats. Collaboration on complex cases is hampered, and research efforts requiring large, harmonized datasets, such as AI training or multi-center trials—become nearly impossible. In practice, data silos mean being excluded from the next wave of precision medicine and big data-driven discovery.

To sum up: compatibility may open the door, but staying there comes at a cost. Over time, it turns into a trap, economically inefficient, clinically risky, and strategically limiting.

Bridging the Gap: A Practical Roadmap

Moving from compatibility to interoperability is not just a technical challenge—it is a cultural and strategic shift. It requires institutions to move away from a device-centered mindset (focused on scanners and viewers) toward a data-centered strategy, where information is the real asset.

Here are some key steps to make that transition real:

1. Adopt open standards from the start.
   When evaluating scanners or viewers, hospitals should require that whole-slide images (WSI) are generated and stored in DICOM format. Likewise, integration with laboratory systems (LIS) and electronic health records (EHRs) should rely on established standards like HL7 or its modern evolution FHIR. This ensures that today’s investment won’t become tomorrow’s barrier.

2. Validate in real-world workflows.
   Pilot projects must go beyond checking image quality. They should test whether data integrates smoothly into actual clinical workflows. Pathologists and technicians should be part of the validation process, ensuring that the solution is not only technically functional but also practical in day-to-day operations. For AI tools, multi-center validation is essential to prove their clinical relevance.

3. Negotiate interoperability in contracts.
   Interoperability should not be left to goodwill—it must be enforced contractually. Procurement contracts should require vendors to comply with open standards and robust security protocols (e.g., AES256 encryption, SHA256 authentication). This reduces dependency on single providers and safeguards patient data.

4. Leverage international initiatives.
   Regulatory bodies are already pushing in this direction. The U.S. FDA promotes medical device interoperability, issuing guidance for manufacturers to embed it in design and pre-market submissions. The College of American Pathologists (CAP) has introduced digital pathology CPT codes, making adoption financially viable. In Europe, programs like the European Health Data Space (EHDS) and Digital Europe are driving cross-border interoperability efforts. Hospitals can align with these frameworks to future-proof their investments.

5. Invest in governance and culture.
   Interoperability is not only about IT—it is about people. Institutions need clear governance structures for data sharing, as well as a cultural shift where pathologists, oncologists, IT teams, and administrators see interoperability as a shared goal. Without this mindset, even the best technical standards will remain underused.

The roadmap is clear: interoperability must be a strategic choice. By shifting focus from short-term fixes to long-term scalability, institutions can unlock the full promise of digital pathology: advanced analytics, AI-driven diagnostics, multicenter collaboration, and personalized medicine.

Reflective Closing

The debate between compatibility and interoperability in digital pathology is more than a technical distinction—it is a statement of vision. Compatibility solves the immediate problem: it opens the door and allows us to see slides on a screen. But it leaves the data isolated, trapped in separate rooms of the hospital.

Interoperability, instead, builds the entire house. It creates a connected infrastructure where pathology data flows together with genomics, radiology, and clinical records—delivering a holistic, precise view of the patient. This is where true transformation happens: when information is no longer fragmented but integrated to serve patients, clinicians, and institutions.

The challenge for healthcare leaders is to recognize that compatibility was a necessary first step, but it cannot be the destination. Investing in interoperability is not a luxury—it is the foundation for advanced research, multicenter collaboration, and the next generation of precision medicine.

The question is not whether digital pathology will become mainstream, but whether institutions will choose to be locked in silos or to lead in a connected future.