FIP Seminar: Using machine learning to optimize how microscopes detect infectious disease
Deep learning algorithms offer a powerful means to automatically analyze the content of medical images. Examples include automated digital microscopic analysis of tissue sections for cancer detection and examining blood smears to diagnose infectious diseases. However, many biological samples of interest are primarily transparent to visible light and contain features that are difficult to resolve with a standard optical microscope. Here, we apply tools from the growing field of deep learning to improve how we image such challenging samples. By merging an optical model of image formation into the pipeline of a convolutional neural network (CNN), we can optimize the physical layout of a digital microscope's illumination and detector. This allows us to capture higher fidelity microscope images that can subsequently be classified with increased accuracy. We demonstrate our joint optimization technique with an experimental microscope that automatically identifies malaria-infected cells with 5-10% higher accuracy than standard and alternative microscope designs.
Roarke Horstmeyer is an assistant professor within Duke's Biomedical Engineering Department. He develops microscopes, cameras and computer algorithms for a wide range of applications, from forming 3D reconstructions of organisms to detecting neural activity deep within tissue. His areas of interest include optics, signal processing, optimization and neuroscience.