Summary:
UCLA researchers in the Department of Electrical Engineering have developed a novel deep learning-based method that performs high-fidelity color image reconstruction using a single hologram.
Background:
Pathology slides are currently the gold standard in diagnostics for many diseases. Accurate color representations of well stained pathology slides are paramount to accurate diagnostics and subsequent clinical care. Current colorization methods in coherent imaging systems produce inaccurate colors and are therefore unacceptable for histology and diagnostic applications. Computational hyperspectral imaging approaches can produce colorization at increased accuracy, but at the cost of complex imaging setup, data acquisition and processing.
Innovation:
A novel deep-neural-network-based method was created to achieve accurate color holographic microscopy. This generative adversarial network (GAN) based technology requires only a single super-resolved hologram acquired under wavelength-multiplexed illumination. The innovation achieves similar performance compared to that of the state-of-the-art absorbance spectrum estimation method, with more than four-fold enhancement in terms of data throughput. The technology was successfully demonstrated on stained lung tissue and prostate tissue sections. This method significantly simplifies the data acquisition procedures, the associated data processing and storage steps, and the imaging hardware required.
Potential Applications:
• Histology
• Other studies that require structural staining of cells or tissues
Advantages:
• Simplified data acquisition
• Simplified data processing
• Simplified data storage steps
• Simplified imaging hardware
• Compatible with coherent microscopy techniques.
Development to Date:
A prototype of the innovation has been built and experimentally demonstrated
Related Papers:
Liu, Tairan, Zhensong Wei, Yair Rivenson, Kevin de Haan, Yibo Zhang, Yichen Wu and Aydogan Ozcan. “Deep learning‐based color holographic microscopy.” Journal of Biophotonics 12 (2019): n. pag.
Reference: UCLA Case No. 2019-737
Lead Inventor: Aydogan Ozcan