2020-795 Single-Shot Autofocusing of Microscopy Images Using Deep Learning (DEEP-R)

Summary:

UCLA researchers in the Department of Electrical Engineering have developed a deep learning-based autofocusing method for microscopy imaging (Deep-R). The method can rapidly and blindly autofocus a single-shot microscopy image of a specimen that is acquired at an arbitrary out-of-focus plane.

Background:

Focus is a critical factor when analyzing an image over an extended spatial or temporal scale. During longitudinal imaging experiments, focus drift due to mechanical thermal fluctuations of the microscope can occur. Algorithmic autofocusing methods have been developed to overcome such drifts but are sensitive to image intensity/contrast and require multi-image capture through an axial scan (search) within the specimen volume, taking time and limiting output image resolution. New methods are needed to rapidly autofocus and capture high-quality microscopic images of specimens, especially for measurements that extend over time covering large fields-of-view.

Innovation:

UCLA researchers in the Department of Electrical Engineering have developed an offline deep-learning based autofocusing method to overcome image drift by rapidly and blindly autofocusing a single-shot microscopy image of a specimen that is acquired at an arbitrary out-of-focus plane. The method has been successfully used with fluorescence and brightfield imaging to autofocus snapshots under different scenarios including a uniform axial defocus as well as a sample tilt within the field-of-view. Deep-R is 15x faster compared to current online algorithmic autofocusing methods and opens up new opportunities for rapid microscopic imaging of large sample areas.