Digital Pathology, Artificial Intelligence & Computational Diagnostics

Transforming cancer diagnosis through innovation

The Department of Pathology & Laboratory Medicine at Roswell Park Comprehensive Cancer Center is advancing the future of cancer care through digital pathology, artificial intelligence (AI) and computational diagnostics. By integrating advanced image analysis, machine learning, genomic data, and clinical information, our teams deliver more precise diagnoses, accelerate turnaround times and support personalized treatment decisions.

Digital pathology enables pathologists to evaluate whole-slide images using high-resolution scanners, while AI-powered tools provide additional insights that may improve diagnostic accuracy, consistency and efficiency.

What is digital pathology?

Digital pathology converts traditional glass slides into high-resolution digital images that can be securely reviewed, shared, and analyzed.

Benefits include:

• Faster access to pathology cases
• Remote consultation and collaboration
• Streamlined workflows
• Enhanced multidisciplinary tumor board review
• Improved archiving and retrieval of cases
• Support for education, research, and clinical trials

Pathologists can review cases from anywhere within the enterprise, facilitating rapid expert consultation across disease-specific teams.

Artificial intelligence in cancer diagnostics

AI serves as an advanced decision-support tool that augments the expertise of pathologists. Applications include:

Tumor detection & classification

AI algorithms can identify suspicious regions within tissue specimens and assist in classifying tumor subtypes.

Quantitative biomarker assessment

Automated image analysis can provide objective measurements of biomarkers such as:

• PD-L1
• HER2
• ER/PR
• Ki-67
• Tumor-infiltrating lymphocytes (TILs)

Tumor microenvironment analysis

Advanced computational tools evaluate interactions between cancer cells, immune cells, and surrounding tissue, providing insights relevant to immunotherapy selection.

Metastasis detection

Machine-learning systems can assist in identifying small metastatic deposits that may be difficult to detect during routine review.

Prognostic modeling

AI-enabled algorithms can integrate pathology findings with molecular and clinical data to help predict disease behavior and treatment response.

Computational pathology

Computational pathology combines:

• Histopathology images
• Molecular diagnostics
• Genomic sequencing
• Radiology data
• Clinical outcomes

Through advanced analytics, researchers can uncover patterns that are invisible to the human eye alone.

Key research areas

Precision oncology

Using pathology and molecular data to identify targeted treatment opportunities.

Biomarker discovery

Developing novel tissue-based and digital biomarkers that predict prognosis and therapeutic response.

Immuno-oncology

Understanding how the immune system interacts with cancer and identifying patients most likely to benefit from immunotherapies.

Predictive analytics

Building models that help forecast recurrence risk, treatment response, and clinical outcomes.

AI-enabled precision medicine

Roswell Park's integrated pathology and molecular diagnostics teams support precision oncology by combining:

• Histologic diagnosis
• Genomic profiling
• Molecular biomarker testing
• Clinical trial eligibility screening
• Treatment-response monitoring

This comprehensive approach helps ensure that patients receive therapies tailored to the biological characteristics of their cancer.

Supporting clinical trials and research

Digital pathology and AI platforms support:

• Translational research programs
• Investigator-initiated clinical trials
• Pharmaceutical collaborations
• Companion diagnostic development
• Multi-institutional research networks

Large-scale image and molecular datasets enable investigators to discover new biomarkers and accelerate cancer research.

Education and innovation

Roswell Park is committed to training the next generation of pathologists and physician-scientists in emerging technologies including:/p>

• Digital pathology training
• AI and machine learning workshops
• Computational pathology research projects
• Fellowship and resident education
• Collaborative research with data scientists and bioinformaticians

Future directions: Building the cancer diagnostics laboratory of the future

Roswell Park is investing in technologies that will shape the next era of pathology:

• Artificial intelligence-assisted diagnosis
• Automated image analytics
• Spatial biology and multiplex imaging
• Digital biomarkers
• Integrated radiology-pathology platforms
• Foundation models for cancer diagnostics
• Predictive and preventive oncology tools

Our vision is to leverage technology to enhance diagnostic excellence, drive scientific discovery, and improve outcomes for every patient.