Back
Science

UCLA Researchers Develop Low-Cost Lensfree Holography Platform for Automated HER2 Scoring in Breast Cancer

View source

UCLA Unveils Low-Cost, AI-Powered Platform for Rapid Breast Cancer Testing

Researchers at the University of California, Los Angeles have developed a new diagnostic platform that combines lensfree holography with deep learning to automate HER2 scoring in breast cancer tissue samples. The findings were recently published in BME Frontiers.

The complete imaging hardware costs less than $980.

A New Approach to Imaging

Unlike traditional microscopes, this platform uses lensfree holography, meaning it requires no objective lenses or mechanical focusing. Instead, it captures holographic diffraction signals from stained tissue sections under RGB laser lighting. This design dramatically simplifies the hardware while maintaining high performance.

The system offers a field of view of 1,250 mm², with an effective imaging throughput of 84 mm² per minute.

How It Achieves Accuracy

To ensure reliable results, the platform employs a sophisticated analysis pipeline:

  • A five-model neural network ensemble processes the captured signals.
  • Bayesian Monte Carlo dropout is used for uncertainty quantification, allowing the system to flag results it is less confident about.

On a blinded dataset of 412 independent tissue samples, the system achieved 84.9% accuracy for four-level HER2 classification and 94.8% accuracy for binary scoring.

Managing Uncertainty

The system is designed to be practical, not just accurate. It successfully filtered out 30.4% of misclassified samples, while only losing 7.2% of correct predictions. This feature is crucial for clinical decision-making, as it helps avoid false results.

Simplified, Cost-Effective Design

For even wider deployment, researchers tested a simplified single blue-light mode. This version maintained decent performance while further reducing hardware costs and complexity.

A Tool for Global Health

The platform’s performance is comparable to high-end brightfield microscopes used in standard digital pathology. However, its true value lies in its accessibility.

The platform aims to provide a practical solution for high-throughput, on-site HER2 testing, particularly in regions lacking advanced medical equipment.