ABSTRACT
Abstract The task of determining for which elements of a random field (e.g., pixels in an image) a certain null hypothesis may be rejected is a relevant problem in several scientific areas. In the current contribution, we introduce a new method for performing this task, the regularized hypothesis testing (RHT) method, focusing on its use in neuroimaging research. RHT is based on the formulation of the hypothesis testing task as a Bayesian estimation problem, with the previous application of a Markovian random field. The latter allows for the incorporation of local spatial information and considers different noise models, including spatially correlated noise. In tests on synthetic data showing regular activation levels on uncorrelated noise fields, RHT furnished a true positive rate (TPR) of 0.97, overcoming the state-of-the-art morphology-based hypothesis testing (MBHT) method and the traditional family-wise error rate (FWER) method, which afforded 0.93 and 0.58, respectively. For fields with highly correlated noise, the TPR provided by RHT was 0.65, and by MBHT and FWER was 0.35 and 0.29, respectively. For tests utilizing real functional magnetic resonance imaging (fMRI) data, RHT managed to locate the activation regions when 60% of the original signal were removed, while MBHT located only one region and FWER located none.
Resumen En varias áreas científicas aparece el problema de determinar los elementos de un campo aleatorio (por ejemplo, píxeles en una imagen) en los que se puede rechazar una cierta hipótesis nula. En este artículo presentamos un nuevo método para realizar esta tarea, centrado en aplicaciones para investigación de neuroimagen. Nuestra propuesta se basa en la formulación de pruebas de hipótesis como un problema de estimación Bayesiana, usando como a priori un campo aleatorio Markoviano, que permite incorporar información espacial local y considera diferentes modelos de ruido, incluido el ruido correlacionado espacialmente. Para pruebas en datos sintéticos con niveles de activación regulares sobre campos de ruido no correlacionado, nuestro método obtiene una tasa de verdaderos positivos (TPR) de 0.97, superando al método del estado del arte MBHT y al método de control FWER que obtienen 0.93 y 0.58 respectivamente; para campos con ruido altamente correlacionado, nuestro método obtiene un TPR de 0.65, mientras que MBHT y FWER obtienen 0.35 y 0.29 respectivamente. Para pruebas con datos reales de fMRI, nuestro método localiza las regiones de activación cuando removemos 60% de la señal original, mientras que MBHT no localiza región alguna y FWER localiza una de las dos regiones.
ABSTRACT
A new framework for phase recovery from a single fringe pattern with closed fringes is proposed. Our algorithm constructs an unwrapped phase from previously computed phases with a simple open-fringe-analysis algorithm, twice applied for analyzing horizontal and vertical oriented fringes, respectively. That reduces the closed-fringe-analysis problem to that of choosing the better phase between the two oriented computed phases and then of estimating the local sign. By propagating the phase sign [and a tilewise constant (DC) term] by regions [here named tiles] instead of a pixelwise phase propagation, our analysis of closed-fringe patterns becomes more robust and faster. Additionally, we propose a multigrid refinement for improving the final computed phase.
