Anderson Winkler, MD, PhD

Research Focus

Dr. Winkler joined UTRGV in 2022. His research program has four active lines of research. The first two are well-established, the third growing, and the fourth is incipient.

Image phenotyping

The work on image phenotyping has improved the assessment and understanding of brain morphology and function, demonstrating, for example, that cortical volume is not an adequate measure of gray matter quantities in the cerebral cortex. This led to the development of better methods for assessing brain cortical surface area and cortical volume. The latter has become a standard in the field.

Additional contributions include the development of a multivariate method for jointly analyzing cortical thickness and area, and contributing to the discovery that resting-state functional brain connectivity is heritable. Ongoing projects involve novel approaches to assess brain myelination and cortical curvature measurements, with applications to Alzheimer’s disease and related dementias.

Statistical methods for imaging data

Work in this area has addressed challenges in the analysis of imaging data, such as multiple testing, spatial statistics, non-independent or repeated-measures data, missing data, and acceleration methods that reduce the need for supercomputing power with large datasets. Results from this research have facilitated the use of exact, non-parametric permutation methods for univariate and multivariate inference, with the development of algorithms that enabled the use of novel multivariate methods in brain imaging.

The work in this area has also led to the discovery of a latent mode of between-subject variation in brain resting-state functional connectivity, and the development of the only valid algorithm for permutation inference in the method known as canonical correlation analysis. Many of these methods have been implemented in PALM (Permutation Analysis of Linear Models), a freely available tool for neuroimaging research.

Psychiatric genetics

Dr. Winkler has been involved with the ENIGMA Consortium since its inception, contributing to the development of its structural processing pipeline. He investigated, together with collaborators, associations between imaging measurements and anxiety diagnoses as defined by the DSM-5, and developed statistical strategies for “mega-analyses” of imaging data from multiple research sites. Ongoing investigations in this area seek to identify genomic variants associated with anxiety using data from multiple research sites worldwide.

Genetics in computational neuroscience

This emerging line of research involves functional magnetic resonance imaging using established study designs to investigate genetic influences on brain processes related to impulsivity, memory, emotion regulation, fear, and pain, as well as pathological processes that involve cognitive decline and dementia. The unique aspect of this line of research lies in its genetic component, and the ability to bring outcomes from discoveries back to the population of the Rio Grande Valley.