Publications

Results 1–25 of 89
Skip to search filters

AirNet-SNL: End-to-end training of iterative reconstruction and deep neural network regularization for sparse-data XPCI CT

Optics InfoBase Conference Papers

Lee, Dennis J.; Mulcahy-Stanislawczyk, Johnathan M.; Jimenez, Edward S.; West, Roger D.; Goodner, Ryan; Epstein, Collin E.; Thompson, Kyle R.; Dagel, Amber L.

We present a deep learning image reconstruction method called AirNet-SNL for sparse view computed tomography. It combines iterative reconstruction and convolutional neural networks with end-to-end training. Our model reduces streak artifacts from filtered back-projection with limited data, and it trains on randomly generated shapes. This work shows promise to generalize learning image reconstruction.

More Details
TYPE Conference Paper YEAR 2021
ScopusOSTIDOI

Influence of Data Acquisition Algorithms on X-Ray Phase Contrast Imaging Computed Tomography

Journal of Nondestructive Evaluation, Diagnostics and Prognostics of Engineering Systems

Epstein, Collin E.; Goodner, Ryan N.; West, Roger D.; Thompson, Kyle R.; Dagel, Amber L.

X-ray phase contrast imaging (XPCI) is a nondestructive evaluation technique that enables high-contrast detection of low-attenuation materials that are largely transparent in traditional radiography. Extending a grating-based Talbot-Lau XPCI system to three-dimensional imaging with computed tomography (CT) imposes two motion requirements: the analyzer grating must translate transverse to the optical axis to capture image sets for XPCI reconstruction, and the sample must rotate to capture angular data for CT reconstruction. The acquisition algorithm choice determines the order of movement and positioning of the two stages. The choice of the image acquisition algorithm for XPCI CT is instrumental to collecting high fidelity data for reconstruction. We investigate how data acquisition influences XPCI CT by comparing two simple data acquisition algorithms and determine that capturing a full phase-stepping image set for a CT projection before rotating the sample results in higher quality data.

More Details
TYPE Journal Article YEAR 2020
ScopusOSTIDOI

Performance evaluation of two optical architectures for task-specific compressive classification

Optical Engineering

Redman, Brian J.; Dagel, Amber L.; Galiardi, Meghan A.; LaCasse, Charles F.; Quach, Tu-Thach Q.; Birch, Gabriel C.

Many optical systems are used for specific tasks such as classification. Of these systems, the majority are designed to maximize image quality for human observers. However, machine learning classification algorithms do not require the same data representation used by humans. We investigate the compressive optical systems optimized for a specific machine sensing task. Two compressive optical architectures are examined: an array of prisms and neutral density filters where each prism and neutral density filter pair realizes one datum from an optimized compressive sensing matrix, and another architecture using conventional optics to image the aperture onto the detector, a prism array to divide the aperture, and a pixelated attenuation mask in the intermediate image plane. We discuss the design, simulation, and trade-offs of these systems built for compressed classification of the Modified National Institute of Standards and Technology dataset. Both architectures achieve classification accuracies within 3% of the optimized sensing matrix for compression ranging from 98.85% to 99.87%. The performance of the systems with 98.85% compression were between an F / 2 and F / 4 imaging system in the presence of noise.

More Details
TYPE Journal Article YEAR 2020
ScopusOSTIDOI

Optimizing a Compressive Imager for Machine Learning Tasks

Conference Record - Asilomar Conference on Signals, Systems and Computers

Redman, Brian J.; Calzada, Daniel; Wingo, Jamie; Quach, Tu-Thach Q.; Galiardi, Meghan; Dagel, Amber L.; LaCasse, Charles F.; Birch, Gabriel C.

Images are often not the optimal data form to perform machine learning tasks such as scene classification. Compressive classification can reduce the size, weight, and power of a system by selecting the minimum information while maximizing classification accuracy.In this work we present designs and simulations of prism arrays which realize sensing matrices using a monolithic element. The sensing matrix is optimized using a neural network architecture to maximize classification accuracy of the MNIST dataset while considering the blurring caused by the size of each prism. Simulated optical hardware performance for a range of prism sizes are reported.

More Details
TYPE Conference Poster YEAR 2019
ScopusOSTIDOI

Double sided grating fabrication for high energy X-ray phase contrast imaging

Materials Science in Semiconductor Processing

Hollowell, Andrew E.; Arrington, Christian L.; Finnegan, Patrick S.; Musick, Katherine M.; Resnick, Paul J.; Volk, Steve; Dagel, Amber L.

State of the art grating fabrication currently limits the maximum source energy that can be used in lab based x-ray phase contrast imaging (XPCI) systems. In order to move to higher source energies, and image high density materials or image through encapsulating barriers, new grating fabrication methods are needed. In this work we have analyzed a new modality for grating fabrication that involves precision alignment of etched gratings on both sides of a substrate, effectively doubling the thickness of the grating. We have achieved a front-to-backside feature alignment accuracy of 0.5 µm demonstrating a methodology that can be applied to any grating fabrication approach extending the attainable aspect ratios allowing higher energy lab based XPCI systems.

More Details
TYPE Journal Article YEAR 2019
ScopusOSTIDOI
Results 1–25 of 89
Results 1–25 of 89