Success Story: Clinical Validation of AI in Dermatopathology with Primaa and PathPresenter
PathPresenter platform enables a smooth study; Primaa CLEO model brings measurable improvements in efficiency and diagnostic agreement
Artificial intelligence is rapidly transforming the field of pathology, but demonstrating its real-world clinical value requires carefully designed validation studies. When AI developer Primaa set out to evaluate the impact of its CLEO dermatopathology model through a clinical study, their team needed a robust platform capable of delivering whole slide images, enabling structured evaluation, and capturing detailed performance metrics from participating pathologists.
PathPresenter was proud to partner with Primaa to support this initiative, providing the secure digital infrastructure that made smooth operation of the study possible.
Evaluating the Impact of AI on Diagnostic Performance
In late 2025, Primaa began a clinical validation study to measure how its CLEO AI model could support pathologists in real-world dermatopathology diagnostic situations. The study compared diagnostic performance when pathologists reviewed cases with and without access to CLEO’s AI-powered assistance.
The goal was twofold: measure improvements in efficiency and assess gains in diagnostic effectiveness.
On the efficiency side, the study evaluated time savings per case and examined how CLEO affected specific diagnostic tasks, including:
- Classification of lesions
- Mitotic counting
- Lesion measurement
For effectiveness, the study focused on key diagnostic quality metrics, including:
- Interobserver agreement among pathologists
- Sensitivity in detecting mitoses
- Accuracy in classifying lesions
Cases included a range of dermatopathology diagnoses such as melanoma, nevus, squamous cell carcinoma (SCC), basal cell carcinoma (BCC), and other benign lesions, providing a realistic clinical mix for evaluation.
To conduct the study successfully, Primaa needed a platform that could provide participating pathologists with seamless access to cases, whole slide images (WSIs), and the CLEO AI model while also capturing detailed interaction data and performance metrics.
A Purpose-Built Platform enabling Clinical Studies
PathPresenter’s Image Management System (IMS) Clinical Module served as the platform for the study, with Primaa’s CLEO model seamlessly integrated into the image viewer. Given both companies’ commitment to interoperability, the integration was quickly and easily done.
Using PathPresenter, participating pathologists were able to access all study materials, including digital slides and associated case information, through a single, intuitive interface that gave easy access to the CLEO model. The platform delivered a consistent experience for each participant, ensuring standardized workflows across the study.
Equally important, the PathPresenter environment provided the security and compliance required for clinical research, meeting strict privacy and data auditability and traceability standards. This allowed Primaa to confidently share study data with participating experts while maintaining full regulatory compliance.
Beyond simple access to slides, PathPresenter enabled the structured collection of study data. The platform captured detailed metrics, including time spent per case and time required for specific diagnostic tasks, allowing the Primaa team to precisely quantify the impact of AI assistance on workflow efficiency.
Because the platform centralized all case review activity, it also made it easier for study organizers to monitor participation, ensure consistency across reviewers, and aggregate results for analysis.
Measurable Improvements in Efficiency and Diagnostic Agreement
The full results of the study are forthcoming, but in the early results from phase 1, the study has already demonstrated positive value for AI-assisted pathology workflows.
Primaa’s CLEO model enabled faster turnaround times per case, helping pathologists complete reviews more efficiently. The study also showed significant improvements in mitotic counting speed, one of the most time-consuming and tedious tasks in dermatopathology evaluation. Overall CLEO enabled a 25% reduction in analysis time.
In addition to efficiency gains, CLEO enabled a 6% improvement in diagnostic accuracy. Pathologists using the AI model demonstrated:
- Increased interobserver agreement when classifying lesions
- Higher sensitivity in detecting mitoses
- Improved accuracy in lesion classification
These findings highlight the potential for AI tools like CLEO to not only accelerate pathology workflows but also enhance diagnostic reliability across practitioners.
Enabling the Future of AI Validation in Pathology
As AI continues to enter clinical practice, rigorous validation studies will play a crucial role in demonstrating safety, effectiveness, and real-world value. Platforms that enable secure collaboration, standardized workflows, and precise measurement are essential to making these studies possible.
PathPresenter is proud to support and partner with organizations like Primaa that are pushing the field forward through transparent, quantitative research.
The full results of Primaa’s CLEO validation study will be available when all phases are complete. More information about the study and Primaa’s work in AI-driven pathology, including skin and breast models, can also be found at primaalab.com.
See it in action at USCAP 2026
Primaa and PathPresenter will demonstrate their solutions at USCAP 2026 in San Antonio, March 23-25. Visit PathPresenter at booth #349 or Primaa at booth #836 for live demonstrations of the full suite of modules and integrations and more on the study results. Interested in a personal demo? Reach out to us at www.pathpresenter.com/contact
More Posts
- Primaa and PathPresenter Introduce Fully-Integrated AI Workflows for Dermatopathology
- The Proof Problem, Part 1: AI Isn’t Replacing Our Job. It’s Adding a New One.
- The Business of Pathology: Unlocking a Lab’s Value
- Expanding Access to Hantavirus Education: Sharing Images Through the PathPresenter Public Library
- The Future of Pathology Part 3: The Delivery Problem and the Infrastructure of Intelligence
