Statistical disease mapping for heterogeneous neuroimaging studies

Abstract

Many cancers and neuro-related diseases display significant phenotypic and genetic heterogeneity across subjects and subpopulations. Characterizing such heterogeneity could transform our understanding of the etiology of these conditions and inspire new approaches to urgently needed prevention, diagnosis, treatment, and prognosis. However, most existing statistical methods face major challenges in delineating such heterogeneity at both the group and individual levels. The aim of this article is to propose a novel statistical disease-mapping (SDM) framework to address some of these challenges. We develop an efficient estimation method to estimate unknown parameters in SDM and delineate individual and group disease maps. Statistical inference procedures such as hypothesis-testing problems are also investigated for parameters of interest. Both simulation studies and real data analysis on the ADNI hippocampal surface dataset show that our SDM not only effectively detects diseased regions in each patient but also provides a group disease-mapping analysis of Alzheimer subgroups.

Presentation of Novel Hybrid Algorithm for Detection and Classification of Breast Cancer Using Growth Region Method and Probabilistic Neural Network

“…According to the l × l transition probability matrix, in which l represents the number of states, the underlying Markov chain alters its state. HMMs have been effectively used in speech recognition and handwritten script recognition [ 52 ]. Since the original HMMs were created as 1D Markov chains, 2D/3D problems like image segmentation cannot be used directly with first-order neighborhood structures.…”

Statistical Learning Methods for Neuroimaging Data Analysis with Applications

“…Developing a good predictive system requires appropriately handling themes T1-T8, among which T4 needs closer attention. Equation 1 emphasizes that neuroimaging data contain external heterogeneity caused by exogenous factors (e.g., the device, acquisition parameters), as well as internal heterogeneity associated with downstream tasks for Y (183). Specifically, "internal heterogeneity" refers to how diseased regions may significantly vary across subjects or time in terms of their numbers, sizes, degrees, and locations.…”

Statistical learning methods for neuroimaging data analysis with applications

“…Developing a good predictive system requires appropriately handling (CT1)-(CT8), among which (CT4) needs more close attention. Model (1) emphasizes that neuroimage data contain external heterogeneity caused by exogenous factors (e.g., device, acquisition parameters) and internal heterogeneity associated with downstream tasks for Y (Liu and Zhu, 2021). Specifically, "internal heterogeneity" refers to how diseased regions may significantly vary across subjects and/or time in terms of their number, size, degree, and location.…”