Deep learning to design imaging hardware

This page is dedicated to deep learning algorithms that design new types of hardware. The primary focus of the work presented here is to showcase new and improved imaging systems (cameras, microscopes, CT, MRI), which are specifically optimized to collect data by and for deep learning tasks. Please find an introduction to this area of research here, and several example projects demonstrating this new effort below.

Duke Computational Optics Lab

This website includes research from the Computational Optics Lab at Duke University. We develop new microscopes, cameras and computer algorithms to capture better biomedical images. The lab is directed by Dr. Roarke Horstmeyer , who is an Assistant Professor in the Department of Biomedical Engineering at Duke

BME Course: Machine Learning and Imaging

The Duke University engineering course, Machine Learning and Imaging, is now fully online. This class is an overview of machine learning and imaging science, with a focus on the intersection of the two fields. It offers much of the introductory material needed to understand the basics of machine learning for hardware design. Please find lectures, homeworks, example code and the course content here


T2oFu: Tensorial tomographic Fourier ptychography with applications to muscle tissue imaging

We report tensorial tomographic Fourier ptychography (T2oFu), a nonscanning label-free tomographic microscopy method for simultaneous imaging of quantitative phase and anisotropic specimen information in 3D. Built upon Fourier ptychography, a quantitative phase imaging technique, T2oFu additionally highlights the vectorial nature of light. We demonstrate T2oFu’s efficiency through volumetric reconstructions of refractive index, birefringence, and orientation for various validation samples, as well as tissue samples from muscle fibers and diseased heart tissue.

Advanced Photonics (2024)

Our review on Digital staining in optical microscopy using deep learning

Over the past few years, digital staining has emerged as a promising concept to use modern deep learning for the translation from optical contrast to established biochemical contrast of actual stainings. In this review article, we provide an in-depth analysis of the current state-of-the-art in this field, suggest methods of good practice, identify pitfalls and challenges and postulate promising advances towards potential future implementations and applications.

PhotoniX (2023)

scatterBrains: an open database of human head models and companion optode locations for realistic Monte Carlo photon simulations

In this work, we provide the NIRS/DCS communities with an open-source, user-friendly database of morphologically and optically realistic head models, as well as a succinct software pipeline to prepare these models for mesh-based Monte Carlo simulations of light transport. The models, head surface locations, and all associated code are freely available under the scatterBrains project on Github. We hope this will make MRI-based head models and virtual optode placement easily accessible to all. Contributions to the database are welcome and encouraged.

Journal of Biomedical Optics (2023)

PDCI: Imaging Dynamics Beneath Turbid Media via Parallelized Single-Photon Detection

In this work, a single-photon avalanche diode array camera is utilized to simultaneously monitor the temporal dynamics of speckle fluctuations at the single-photon level from 12 different phantom tissue surface locations delivered via a customized fiber bundle array. Then a deep neural network is applied to convert the acquired single-photon measurements into video of scattering dynamics beneath rapidly decorrelating tissue phantoms. The ability to reconstruct images of transient (0.1–0.4 s) dynamic events occurring up to 8 mm beneath a decorrelating tissue phantom with millimeter-scale resolution is demonstrated, and it is highlighted how the model can flexibly extend to monitor flow speed within buried phantom vessels.

Advanced Science (2022)

Vectorial FP

We present a microscopic imaging technique that uses variable-angle illumination to recover the complex polarimetric properties of a specimen at high resolution and over a large field-of-view. The approach extends Fourier ptychography, to additionally account for the vectorial nature of light.

Biomedical Optics Express (2022)

Virtual Fluoroscence

Our results show that by including a model of illumination within the first layers of a deep convolutional neural network, it is possible to learn task-specific LED patterns that substantially improve the ability to infer fluorescence image information from unstained transmission microscopy images.

ICCV (2021)


We study the feasibility of implementing Fourier ptychography (FP) with SPAD array cameras to reconstruct an image with higher resolution and larger dynamic range from acquired binary intensity measurements.

Photonics Research (2021)

Deep Optical Blood Analysis

We propose a novel machine learning approach that offers a powerful and scalable means to analyse peripheral blood smears by utilizing a multiple instance learning based approach to understand the morphological impact of COVID-19 on the blood cells across the various cell types, which is still not well understood.

medRxiv (2021)

Parallelized Diffuse Correlation Spectrocopy

With a high-sensitivity DCS system with 1024 parallel detection channels integrated within a SPAD array, we demonstrate the ability to detect mm-scale perturbations up to 1 cm deep within a tissue-like phantom . We also measure the human pulse at high fidelity and detect behaviorally-induced physiological variations from the human prefrontal cortex.

APL Photonics (2020)

Deep Prior Diffraction Tomography

We present a tomographic imaging technique, termed Deep Prior Diffraction Tomography (DP-DT), to reconstruct the 3D refractive index (RI) of thick biological samples at high resolution from a sequence of low-resolution images collected under angularly varying illumination.

Optics Express (2020)

Multi Element microscope optimization

In this work, we investigate an approach to jointly optimize multiple microscope settings, together with a classification network, for improved performance with such automated tasks.

Optics Letters (2020)

Towards an intelligent microscope

Illumination within a microscope can drastically alter information captured by the image sensor. We present a reinforcement learning system that adaptively explores optimal patterns to illuminate specimens for immediate classification.

ICASSP (2020)

Learned Sensing

To increase the speed and accuracy of automated image classification, this work presents a method to co-optimize how a sample is illuminated in a microscope, along with a pipeline to automatically classify the resulting image, using a deep neural network

Biomedical Optics Express (2019)

MCAM: The Gigapixel Microscope

We present a novel microscopy approach: Multi Camera Array Microscopy (MCAM), for rapid acquisition of Gigapixel Videos and Images. Also capture 3D Depth measurement, and analyse large FOV sample such as 48 well plates and live freely moving model organisms, with inbuilt algorithms

In preparation…

// First try loading jQuery from Google's CDN