BlurryScope: An Affordable, AI-Driven Microscope Revolutionizing Cancer Diagnostics

A pioneering research group at UCLA, led by Professor Aydogan Ozcan, has developed BlurryScope, a revolutionary, compact microscope that leverages artificial intelligence and simple optical hardware to enhance cancer diagnosis, specifically HER2 status in breast tissue. Unlike traditional high-cost and bulky digital pathology scanners, BlurryScope can be constructed for under $650, measuring only 35 x 35 x 35 centimeters and weighing just 2.26 kilograms, making it highly portable and accessible.

The core innovation of BlurryScope lies in its unique image capture technique. Instead of acquiring high-resolution, still images that require precise mechanical control and longer scan times, it continuously scans tissue samples, producing motion-blurred images, typically considered unusable. An advanced deep learning model is trained to interpret these blurred images accurately, classifying HER2 expression levels—a critical biomarker for guiding breast cancer treatment.

Published in npj Digital Medicine, the study illustrates how AI can extract significant diagnostic information from images that would traditionally be disregarded, reducing both the complexity and expense of digital pathology devices.

In tests involving 284 patient tissue samples, BlurryScope achieved nearly 80% accuracy across four HER2 scoring categories. When scores were grouped into two clinically relevant categories, accuracy increased to nearly 90%. The system demonstrated high reproducibility, with over 86% consistency across multiple scans, indicating its robustness for clinical application.

This automation streamlines the diagnostic process, from continuous slide scanning to image analysis, making it an effective tool for rapid screening, especially in settings where traditional high-end equipment is unaffordable or impractical. The technology exemplifies how integrating computational methods with optical systems can significantly reduce hardware requirements while maintaining high diagnostic standards.

Beyond HER2 assessment, the principles underlying BlurryScope could extend to other tissue biomarkers or different clinical imaging scenarios, underscoring a new paradigm in biomedical engineering that combines optical innovation with AI analytics. Dr. Ozcan emphasizes that this approach could democratize access to quality diagnostics worldwide, ultimately transforming cancer care by making it more affordable and accessible.

For more detailed information, refer to the full study: npj Digital Medicine.

Source: https://medicalxpress.com/news/2025-09-blurryscope-compact-ai-powered-microscope.html