-
Heterogeneous Image-based Classification Using Distributional Data Analysis
Authors:
Alec Reinhardt,
Newsha Nikzad,
Raven J. Hollis,
Galia Jacobson,
Millicent A. Roach,
Mohamed Badawy,
Peter Chul Park,
Laura Beretta,
Prasun K Jalal,
David T. Fuentes,
Eugene J. Koay,
Suprateek Kundu
Abstract:
Diagnostic imaging has gained prominence as potential biomarkers for early detection and diagnosis in a diverse array of disorders including cancer. However, existing methods routinely face challenges arising from various factors such as image heterogeneity. We develop a novel imaging-based distributional data analysis (DDA) approach that incorporates the probability (quantile) distribution of the…
▽ More
Diagnostic imaging has gained prominence as potential biomarkers for early detection and diagnosis in a diverse array of disorders including cancer. However, existing methods routinely face challenges arising from various factors such as image heterogeneity. We develop a novel imaging-based distributional data analysis (DDA) approach that incorporates the probability (quantile) distribution of the pixel-level features as covariates. The proposed approach uses a smoothed quantile distribution (via a suitable basis representation) as functional predictors in a scalar-on-functional quantile regression model. Some distinctive features of the proposed approach include the ability to: (i) account for heterogeneity within the image; (ii) incorporate granular information spanning the entire distribution; and (iii) tackle variability in image sizes for unregistered images in cancer applications. Our primary goal is risk prediction in Hepatocellular carcinoma that is achieved via predicting the change in tumor grades at post-diagnostic visits using pre-diagnostic enhancement pattern map** (EPM) images of the liver. Along the way, the proposed DDA approach is also used for case versus control diagnosis and risk stratification objectives. Our analysis reveals that when coupled with global structural radiomics features derived from the corresponding T1-MRI scans, the proposed smoothed quantile distributions derived from EPM images showed considerable improvements in sensitivity and comparable specificity in contrast to classification based on routinely used summary measures that do not account for image heterogeneity. Given that there are limited predictive modeling approaches based on heterogeneous images in cancer, the proposed method is expected to provide considerable advantages in image-based early detection and risk prediction.
△ Less
Submitted 11 March, 2024;
originally announced March 2024.
-
The Concept of Force in Population Dynamics
Authors:
John Hayward,
Paul A. Roach
Abstract:
The area of population dynamics has a rich history of the development and analysis of models of biological and social phenomena using ordinary differential equations. This paper describes a method for understanding the influence one variable exerts on another in such models as a force, with the relative effects of these forces providing a narrative explanation of the curvature in variable behaviou…
▽ More
The area of population dynamics has a rich history of the development and analysis of models of biological and social phenomena using ordinary differential equations. This paper describes a method for understanding the influence one variable exerts on another in such models as a force, with the relative effects of these forces providing a narrative explanation of the curvature in variable behaviour. Using the stock/flow form of a model, a symbolic notation is developed that identifies the forces with the causal pathways of the model's feedback loops. A force is measured by its impact, defined as the ratio of acceleration to rate of change, computed by differentiation along its associated pathway between variables. Different phases of force dominance are determined to enhance the standard stability analysis of the models, providing an explanation of model behaviour in Newtonian mechanical terms. The concepts developed are applied to well-known models from mathematical biology: the Spruce Budworm model, where force dominance identifies scenarios that give clarity to intervention points; and the Lotka-Volterra predator-prey model where the analysis highlights the importance of dissipative forces in achieving stability. Conclusions are drawn on the explanatory power of this approach, with suggestions made for future work.
△ Less
Submitted 8 March, 2022;
originally announced March 2022.
-
Physical drivers of ocean wave attenuation in the marginal ice zone
Authors:
Fabien Montiel,
Alison L. Kohout,
Lettie A. Roach
Abstract:
Despite a recent resurgence of observational studies attempting to quantify the ice-induced attenuation of ocean waves in polar oceans, the physical processes governing this wave attenuation phenomenon are still poorly understood. Most analyses have attempted to relate the spatial rate of wave attenuation to wave frequency, but have not considered how this relationship depends on ice, wave and atm…
▽ More
Despite a recent resurgence of observational studies attempting to quantify the ice-induced attenuation of ocean waves in polar oceans, the physical processes governing this wave attenuation phenomenon are still poorly understood. Most analyses have attempted to relate the spatial rate of wave attenuation to wave frequency, but have not considered how this relationship depends on ice, wave and atmospheric conditions. An in-depth analysis of the wave-buoy data collected during the 2017 PIPERS programme in the Ross Sea is conducted. Standard techniques are used to estimate the spatial rate of wave attenuation $α$ and the influence of a number of potential physical drivers on its dependence on wave period $T$ is investigated. A power-law is shown to consistently describe the $α(T)$ relationship, in line with other recent analyses. The two parameters describing this relationship are found to depend significantly on sea ice concentration, mean wave period and wind direction, however. Looking at cross-correlations between these physical drivers, three regimes of ice-induced wave attenuation are identified, which characterise different ice, wave and wind conditions, and very possibly different processes causing this observed attenuation. This analysis suggests that parametrisations of ice-induced wave decay in spectral wave models should be piecewise, so their dependence on local ice, wave and wind conditions is described in some way.
△ Less
Submitted 8 November, 2021;
originally announced November 2021.
-
Bishop Independence on the Surface of a Square Prism
Authors:
Liam H. Harris,
Stephanie Perkins,
Paul A. Roach
Abstract:
Bishop independence concerns determining the maximum number of bishops that can be placed on a board such that no bishop can attack any other bishop. This paper presents the solution to the bishop independence problem, determining the bishop independence number, for all sizes of boards on the surface of a square prism.
Bishop independence concerns determining the maximum number of bishops that can be placed on a board such that no bishop can attack any other bishop. This paper presents the solution to the bishop independence problem, determining the bishop independence number, for all sizes of boards on the surface of a square prism.
△ Less
Submitted 29 June, 2020;
originally announced June 2020.
-
Activist Model of Political Party Growth
Authors:
Rebecca A. Jeffs,
John Hayward,
Paul A. Roach,
John Wyburn
Abstract:
The membership of British political parties has a direct influence on their political effectiveness. This paper applies the mathematics of epidemiology to the analysis of the growth and decline of such memberships. The party members are divided into activists and inactive members, where all activists influence the quality of party recruitment, but only a subset of activists recruit and thus govern…
▽ More
The membership of British political parties has a direct influence on their political effectiveness. This paper applies the mathematics of epidemiology to the analysis of the growth and decline of such memberships. The party members are divided into activists and inactive members, where all activists influence the quality of party recruitment, but only a subset of activists recruit and thus govern numerical growth. The activists recruit for only a limited period, which acts as a restriction on further party growth. This Limited Activist model is applied to post-war and recent memberships of the Labour, Scottish National and Conservative parties. The model reproduces data trends, and relates realistically to historical narratives. It is concluded that the political parties analysed are not in danger of extinction but experience repeated periods of growth and decline in membership, albeit at lower numbers than in the past.
△ Less
Submitted 25 September, 2015;
originally announced September 2015.
-
Observation of a Free-Shercliff-Layer Instability in Cylindrical Geometry
Authors:
Austin Roach,
Erik Spence,
Christophe Gissinger,
Eric Edlund,
Peter Sloboda,
Jeremy Goodman,
Hantao Ji
Abstract:
We report on observations of a free-Shercliff-layer instability in a Taylor-Couette experiment using a liquid metal over a wide range of Reynolds numbers, $Re\sim 10^3-10^6$. The free Shercliff layer is formed by imposing a sufficiently strong axial magnetic field across a pair of differentially rotating axial endcap rings. This layer is destabilized by a hydrodynamic Kelvin-Helmholtz-type instabi…
▽ More
We report on observations of a free-Shercliff-layer instability in a Taylor-Couette experiment using a liquid metal over a wide range of Reynolds numbers, $Re\sim 10^3-10^6$. The free Shercliff layer is formed by imposing a sufficiently strong axial magnetic field across a pair of differentially rotating axial endcap rings. This layer is destabilized by a hydrodynamic Kelvin-Helmholtz-type instability, characterized by velocity fluctuations in the $r-θ$ plane. The instability appears with an Elsasser number above unity, and saturates with an azimuthal mode number $m$ which increases with the Elsasser number. Measurements of the structure agree well with 2D global linear mode analyses and 3D global nonlinear simulations. These observations have implications for a range of rotating MHD systems in which similar shear layers may be produced.
△ Less
Submitted 27 April, 2012;
originally announced April 2012.
-
Observation of magnetocoriolis waves in a liquid metal Taylor-Couette experiment
Authors:
M. D. Nornberg,
H. Ji,
E. Schartman,
A. Roach,
J. Goodman
Abstract:
The first observation of fast and slow magnetocoriolis (MC) waves in a laboratory experiment is reported. Rotating nonaxisymmetric modes arising from a magnetized turbulent Taylor-Couette flow of liquid metal are identified as the fast and slow MC waves by the dependence of the rotation frequency on the applied field strength. The observed slow MC wave is damped but the observation provides a me…
▽ More
The first observation of fast and slow magnetocoriolis (MC) waves in a laboratory experiment is reported. Rotating nonaxisymmetric modes arising from a magnetized turbulent Taylor-Couette flow of liquid metal are identified as the fast and slow MC waves by the dependence of the rotation frequency on the applied field strength. The observed slow MC wave is damped but the observation provides a means for predicting the onset of the Magnetorotational Instability.
△ Less
Submitted 19 February, 2010;
originally announced February 2010.
-
The Medieval Inquisition: Scale-free Networks and the Suppression of Heresy
Authors:
Paul Ormerod,
Andrew Roach
Abstract:
Qualitative evidence suggests that heresy within the medieval Catholic Church had many of the characteristics of a scale-free network. From the perspective of the Church, heresy can be seen as a virus. The virus persisted for long periods of time, breaking out again even when the Church believed it to have been eradicated. A principal mechanism of heresy was through a small number of individuals…
▽ More
Qualitative evidence suggests that heresy within the medieval Catholic Church had many of the characteristics of a scale-free network. From the perspective of the Church, heresy can be seen as a virus. The virus persisted for long periods of time, breaking out again even when the Church believed it to have been eradicated. A principal mechanism of heresy was through a small number of individuals with very large numbers of social contacts.
Initial attempts by the Inquisition to suppress the virus by general persecution, or even mass slaughtering, of populations thought to harbour the "disease" failed. Gradually, however, the Inquisition learned about the nature of the social networks by which heresy both spread and persisted. Eventually, a policy of targeting key individuals was implemented, which proved to be much more successful.
△ Less
Submitted 2 June, 2003;
originally announced June 2003.
