Showing 1–8 of 8 results for author: O'Connor, D

Data Quality Over Quantity: Pitfalls and Guidelines for Process Analytics

Authors: Lim C. Siang, Shams Elnawawi, Lee D. Rippon, Daniel L. O'Connor, R. Bhushan Gopaluni

Abstract: A significant portion of the effort involved in advanced process control, process analytics, and machine learning involves acquiring and preparing data. Literature often emphasizes increasingly complex modelling techniques with incremental performance improvements. However, when industrial case studies are published they often lack important details on data acquisition and preparation. Although da… ▽ More A significant portion of the effort involved in advanced process control, process analytics, and machine learning involves acquiring and preparing data. Literature often emphasizes increasingly complex modelling techniques with incremental performance improvements. However, when industrial case studies are published they often lack important details on data acquisition and preparation. Although data pre-processing is unfairly maligned as trivial and technically uninteresting, in practice it has an out-sized influence on the success of real-world artificial intelligence applications. This work describes best practices for acquiring and preparing operating data to pursue data-driven modelling and control opportunities in industrial processes. We present practical considerations for pre-processing industrial time series data to inform the efficient development of reliable soft sensors that provide valuable process insights. △ Less

Submitted 5 April, 2023; v1 submitted 11 November, 2022; originally announced November 2022.

Comments: This work has been accepted to the 22nd IFAC World Congress 2023

A Simple Discretization Scheme for Gain Matrix Conditioning

Authors: Daniel L. O'Connor, Lim C. Siang, Shams Elnawawi

Abstract: In industrial model predictive controllers (MPCs), models generated from regression-based system identification methods typically contain small or even physically non-existent degrees of freedom. Control issues can arise when the steady-state optimizer uses these small degrees of freedom to calculate targets for plant operation due to matrix ill-conditioning. Mathematical techniques like Relative… ▽ More In industrial model predictive controllers (MPCs), models generated from regression-based system identification methods typically contain small or even physically non-existent degrees of freedom. Control issues can arise when the steady-state optimizer uses these small degrees of freedom to calculate targets for plant operation due to matrix ill-conditioning. Mathematical techniques like Relative Gain Array (RGA) and Singular Value Decomposition (SVD) are helpful for analyzing controller gain interactions and identifying conditioning issues, which can be corrected relatively easily in small models. However, these techniques are difficult and tedious to apply for larger, more complex models. This paper describes a novel, non-iterative, RGA-based, binning technique for discretizing the gain matrix and quickly solving 2x2 conditioning issues for any model size, while guaranteeing gain adjustments below a certain threshold. Higher order interactions are also discussed. △ Less

Submitted 9 June, 2022; v1 submitted 26 February, 2022; originally announced February 2022.

Comments: Revised manuscript accepted in AdCONIP 2022

Least Squares Optimal Density Compensation for the Gridding Non-uniform Discrete Fourier Transform

Authors: Nicholas Dwork, Daniel O'Connor, Ethan M. I. Johnson, Corey A. Baron, Jeremy W. Gordon, John M. Pauly, Peder E. Z. Larson

Abstract: The Gridding algorithm has shown great utility for reconstructing images from non-uniformly spaced samples in the Fourier domain in several imaging modalities. Due to the non-uniform spacing, some correction for the variable density of the samples must be made. Existing methods for generating density compensation values are either sub-optimal or only consider a finite set of points (a set of measu… ▽ More The Gridding algorithm has shown great utility for reconstructing images from non-uniformly spaced samples in the Fourier domain in several imaging modalities. Due to the non-uniform spacing, some correction for the variable density of the samples must be made. Existing methods for generating density compensation values are either sub-optimal or only consider a finite set of points (a set of measure 0) in the optimization. This manuscript presents the first density compensation algorithm for a general trajectory that takes into account the point spread function over a set of non-zero measure. We show that the images reconstructed with Gridding using the density compensation values of this method are of superior quality when compared to density compensation weights determined in other ways. Results are shown with a numerical phantom and with magnetic resonance images of the abdomen and the knee. △ Less

Submitted 16 June, 2021; v1 submitted 11 June, 2021; originally announced June 2021.

Slide-free MUSE Microscopy to H&E Histology Modality Conversion via Unpaired Image-to-Image Translation GAN Models

Authors: Tanishq Abraham, Andrew Shaw, Daniel O'Connor, Austin Todd, Richard Levenson

Abstract: MUSE is a novel slide-free imaging technique for histological examination of tissues that can serve as an alternative to traditional histology. In order to bridge the gap between MUSE and traditional histology, we aim to convert MUSE images to resemble authentic hematoxylin- and eosin-stained (H&E) images. We evaluated four models: a non-machine-learning-based color-map** unmixing-based tool, Cy… ▽ More MUSE is a novel slide-free imaging technique for histological examination of tissues that can serve as an alternative to traditional histology. In order to bridge the gap between MUSE and traditional histology, we aim to convert MUSE images to resemble authentic hematoxylin- and eosin-stained (H&E) images. We evaluated four models: a non-machine-learning-based color-map** unmixing-based tool, CycleGAN, DualGAN, and GANILLA. CycleGAN and GANILLA provided visually compelling results that appropriately transferred H&E style and preserved MUSE content. Based on training an automated critic on real and generated H&E images, we determined that CycleGAN demonstrated the best performance. We have also found that MUSE color inversion may be a necessary step for accurate modality conversion to H&E. We believe that our MUSE-to-H&E model can help improve adoption of novel slide-free methods by bridging a perceptual gap between MUSE imaging and traditional histology. △ Less

Submitted 19 August, 2020; originally announced August 2020.

Comments: 4 pages plus 1 page references. Presented at the ICML Computational Biology Workshop 2020

Multi-coil Magnetic Resonance Imaging with Compressed Sensing Using Physically Motivated Regularization

Authors: Nicholas Dwork, Ethan M. I. Johnson, Daniel O'Connor, Jeremy W. Gordon, Adam B. Kerr, Corey A. Baron, John M. Pauly, Peder E. Z. Larson

Abstract: With the advent of multi-coil imaging and compressed sensing, a number of model based reconstruction algorithms have been created. They incorporate a multitude of different regularization functions based on physics, observed phenomenology, and heuristics. Moreover, several iterative methods exist that attempt to simultaneously estimate the sensitivity maps and the image. In this manuscript, we pre… ▽ More With the advent of multi-coil imaging and compressed sensing, a number of model based reconstruction algorithms have been created. They incorporate a multitude of different regularization functions based on physics, observed phenomenology, and heuristics. Moreover, several iterative methods exist that attempt to simultaneously estimate the sensitivity maps and the image. In this manuscript, we present a generalization of several existing iterative model based algorithms. We devise a calibrationless instance of this generalization that only incorporates regularization terms based on physics and the accepted compressed sensing phenomenology of sparsity in the wavelet domain. We compare the results of the new amalgamated optimization problem with existing methods on both simulated and real datasets. We show that the images reconstructed using the new method, entitled Multi-coil Compressed Sensing (MCCS), are of higher quality than existing methods in all cases studied. △ Less

Submitted 2 February, 2023; v1 submitted 30 June, 2020; originally announced July 2020.

Fast Variable Density Poisson-Disc Sample Generation with Directional Variation

Authors: Nicholas Dwork, Corey A. Baron, Ethan M. I. Johnson, Daniel O'Connor, John M. Pauly, Peder E. Z. Larson

Abstract: We present a fast method for generating random samples according to a variable density Poisson-disc distribution. A minimum threshold distance is used to create a background grid array for kee** track of those points that might affect any new candidate point; this reduces the number of conflicts that must be checked before acceptance of a new point, thus reducing the number of computations requi… ▽ More We present a fast method for generating random samples according to a variable density Poisson-disc distribution. A minimum threshold distance is used to create a background grid array for kee** track of those points that might affect any new candidate point; this reduces the number of conflicts that must be checked before acceptance of a new point, thus reducing the number of computations required. We demonstrate the algorithm's ability to generate variable density Poisson-disc sampling patterns according to a parameterized function, including patterns where the variations in density are a function of direction. We further show that these sampling patterns are appropriate for compressed sensing applications. Finally, we present a method to generate patterns with a specific acceleration rate. △ Less

Submitted 16 June, 2021; v1 submitted 14 April, 2020; originally announced April 2020.

Journal ref: Magnetic Resonance Imaging 77 (2021): 186-193

Di-chromatic Interpolation of Magnetic Resonance Metabolic Imagery

Authors: Nicholas Dwork, Jeremy W. Gordon, Shuyu Tang, Daniel O'Connor, Esben Sovso Szocska Hansen, Christoffer Laustsen, Peder E. Z. Larson

Abstract: Magnetic resonance imaging with hyperpolarized contrast agents can provide unprecedented \textit{in-vivo} measurements of metabolism, but yields images that are lower resolution than that achieved with proton anatomical imaging. In order to spatially localize the metabolic activity, the metabolic image must be interpolated to the size of the proton image. The most common methods for choosing the u… ▽ More Magnetic resonance imaging with hyperpolarized contrast agents can provide unprecedented \textit{in-vivo} measurements of metabolism, but yields images that are lower resolution than that achieved with proton anatomical imaging. In order to spatially localize the metabolic activity, the metabolic image must be interpolated to the size of the proton image. The most common methods for choosing the unknown values rely exclusively on values of the original un-interpolated image. In this work, we present an alternative method that uses the higher-resolution proton image to provide additional spatial structure. The interpolated image is the result of a convex optimization algorithm which is solved with the Fast Iterative Shrinkage Threshold Algorithm (FISTA). Results are shown with images of hyperpolarized pyruvate, lactate, and bicarbonate using data of the heart and brain from healthy human volunteers, a healthy porcine heart, and a human with prostate cancer. △ Less

Submitted 12 February, 2021; v1 submitted 10 March, 2020; originally announced March 2020.

Comments: Magnetic Resonance Materials in Physics, Biology, and Medicine (2021)

Journal ref: Magnetic Resonance Materials in Physics, Biology and Medicine (2021): 1-16

Utilizing the Wavelet Transform's Structure in Compressed Sensing

Authors: Nicholas Dwork, Daniel O'Connor, Corey A. Baron, Ethan M. I. Johnson, Adam B. Kerr, John M. Pauly, Peder E. Z. Larson

Abstract: Compressed sensing has empowered quality image reconstruction with fewer data samples than previously though possible. These techniques rely on a sparsifying linear transformation. The Daubechies wavelet transform is a common sparsifying transformation used for this purpose. In this work, we take advantage of the structure of this wavelet transform and identify an affine transformation that increa… ▽ More Compressed sensing has empowered quality image reconstruction with fewer data samples than previously though possible. These techniques rely on a sparsifying linear transformation. The Daubechies wavelet transform is a common sparsifying transformation used for this purpose. In this work, we take advantage of the structure of this wavelet transform and identify an affine transformation that increases the sparsity of the result. After inclusion of this affine transformation, we modify the resulting optimization problem to comply with the form of the Basis Pursuit Denoising problem. Finally, we show theoretically that this yields a lower bound on the error of the reconstruction and present results where solving this modified problem yields images of higher quality for the same sampling patterns. △ Less

Submitted 16 June, 2021; v1 submitted 10 February, 2020; originally announced February 2020.

Journal ref: Signal, Image and Video Processing (2021): 1-8