How PathAI’s PathAssist Derm Tool Aims to Enhance Skin Cancer Diagnosis and Workflow Efficiency

In an interview with Dermatology Times, Andy Beck, MD, PhD, CEO and co-founder of PathAI, discussed the development of the PathAssist Derm tool, which was designed to assist dermatopathologists in diagnosing skin malignancies. Beck explained that dermatopathology plays a crucial role in diagnosing skin diseases, but a shortage of specialized dermatopathologists makes accurate diagnosis challenging. To address this, PathAssist Derm was developed to predict diagnoses for 17 skin malignancies from actinic keratoses and basal cell carcinomas to lichenoid keratoses and melanomas. The tool’s performance was validated across a wide range of conditions, from common to rare skin diseases, demonstrating strong diagnostic accuracy.

Beck emphasized the difficulty of creating a tool for such a broad spectrum of dermatologic conditions, including both inflammatory and neoplastic diseases, which can present with varying appearances. PathAssist Derm was built using a large machine-learning model trained on millions of labeled image parts and data from hundreds of thousands of patients. This extensive dataset formed the foundation for the tool, which was further refined with tens of thousands of dermatopathology slides to ensure its ability to handle a broad array of diagnoses.

“Putting together the right data set for both training and validating and testing the performance of the system was a major part of the work we did for building this PathAssist Derm tool,” Beck said.

Looking ahead, Beck discussed the integration of PathAssist Derm into clinical practice. Currently, the tool is designed for research use only, as it reduces the manual assessment time while improving efficiency and diagnostic precision. However, PathAI is conducting pilot studies with early partners to assess the tool's accuracy and potential clinical impact. Ultimately, Beck envisions the tool’s use in clinical settings, aiming to enhance workflows by automating processes like specimen orientation, prioritization of urgent cases, and measurement tasks—functions that are essential for dermatopathologists in both research and clinical environments. The company is working toward FDA clearance, with plans to validate the clinical version further in the future.

PathAssist Derm is available on the AISight® Image Management System. Further performance data will be shared at the United States and Canadian Academy of Pathology’s Annual Meeting in March.