Xinyu Zhang
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Research Interests
Xinyu Zhang has a total of 48 co-authors. Here are the top co-authors by number of shared publications:
- Simon Vandekar
Department of Biostatistics, Vanderbilt University, Nashville, TN, United States
3 shared publications Aaron Alexander-Bloch
Department of Psychiatry, University of Pennsylvania, Philadelphia, PA, United States
2 shared publicationsAnna Huang
Department of Psychiatry and Behavioral Sciences, Vanderbilt University Medical Center, Nashville, TN, United States
2 shared publicationsJakob Seidlitz
Department of Psychiatry, University of Pennsylvania, Philadelphia, PA, United States
2 shared publicationsKenneth Liao
Department of Biostatistics, Vanderbilt University, Nashville, TN, United States
2 shared publicationsMaureen McHugo
Department of Psychiatry, University of Colorado Anschutz Medical Campus, Aurora, CO, United States
2 shared publications
Latest Publications
- Open Access
- 10 November 2026
- Published by MIT Press in Imaging Neuroscience
- Vol. 4
Optimized multi-echo gradient echo MRI of the lumbosacral spinal cord to investigate atrophy, lesions, and disability in relapsing-remitting multiple sclerosis
The lumbosacral enlargement (LSE) of the spinal cord is understudied in people with multiple sclerosis (pwMS), and there is not a consensus on the impact of age, sex, height, or weight on MRI-derived morphometrics in this region. The LSE is relevant to study in pwMS as it is involved in the control of motor and sensory functions of the lower limbs and autonomic functions which are commonly affected in MS. This study aimed to characterize biological variability of morphometric measures of the LSE cord and gray matter in healthy controls (HCs), compare these measures between HCs and people with relapsing-remitting MS (pwRRMS), and assess correlations between morphometrics and lower-extremity disability and bladder symptoms in pwRRMS. An optimized T2*-weighted, axial multi-echo gradient echo anatomical MRI scan was acquired at 3T, images were automatically segmented, and whole-cord and gray matter cross-sectional areas for the lumbosacral spinal levels of 60 HCs and 42 pwRRMS were computed. We observed no significant main effects of age, height, weight, or sex on LSE whole-cord or gray matter cross-sectional areas (CSA) in HCs, but there were spinal level interactions with age and weight. We detected whole-cord and gray matter atrophy in the LSE of pwRRMS compared to HCs, and the effects of disease status on these measures were partially mitigated when controlling for MS lesion to cord CSA ratio within the LSE. There were no significant associations between LSE cross-sectional area measures and sensorimotor function, bladder symptoms, or clinical disability scores. However, vibration sensation in the great toe and Expanded Disability Status Scale (EDSS) scores were significantly associated with lesion to cord CSA ratio. Overall, this study found that there is little variability linked to age, sex, height, or weight in cross-sectional area measures for the healthy lumbosacral spinal cord, and lesions and atrophy were both detected in these spinal levels for a pwRRMS cohort. Although there were no associations between atrophy and symptoms, potentially due to the relatively low level of clinical disability in this cohort, these findings support the inclusion of the LSE in future imaging studies that aim to probe the full extent of spinal cord pathology in MS.
- 10 November 2026
- Published by MIT Press in Imaging Neuroscience
- Vol. 4
Open AccessOptimized multi-echo gradient echo MRI of the lumbosacral spinal cord to investigate atrophy, lesions, and disability in relapsing-remitting multiple sclerosis
The lumbosacral enlargement (LSE) of the spinal cord is understudied in people with multiple sclerosis (pwMS), and there is not a consensus on the impact of age, sex, height, or weight on MRI-derived morphometrics in this region. The LSE is relevant to study in pwMS as it is involved in the control of motor and sensory functions of the lower limbs and autonomic functions which are commonly affected in MS. This study aimed to characterize biological variability of morphometric measures of the LSE cord and gray matter in healthy controls (HCs), compare these measures between HCs and people with relapsing-remitting MS (pwRRMS), and assess correlations between morphometrics and lower-extremity disability and bladder symptoms in pwRRMS. An optimized T2*-weighted, axial multi-echo gradient echo anatomical MRI scan was acquired at 3T, images were automatically segmented, and whole-cord and gray matter cross-sectional areas for the lumbosacral spinal levels of 60 HCs and 42 pwRRMS were computed. We observed no significant main effects of age, height, weight, or sex on LSE whole-cord or gray matter cross-sectional areas (CSA) in HCs, but there were spinal level interactions with age and weight. We detected whole-cord and gray matter atrophy in the LSE of pwRRMS compared to HCs, and the effects of disease status on these measures were partially mitigated when controlling for MS lesion to cord CSA ratio within the LSE. There were no significant associations between LSE cross-sectional area measures and sensorimotor function, bladder symptoms, or clinical disability scores. However, vibration sensation in the great toe and Expanded Disability Status Scale (EDSS) scores were significantly associated with lesion to cord CSA ratio. Overall, this study found that there is little variability linked to age, sex, height, or weight in cross-sectional area measures for the healthy lumbosacral spinal cord, and lesions and atrophy were both detected in these spinal levels for a pwRRMS cohort. Although there were no associations between atrophy and symptoms, potentially due to the relatively low level of clinical disability in this cohort, these findings support the inclusion of the LSE in future imaging studies that aim to probe the full extent of spinal cord pathology in MS.
- Open Access
- 23 January 2026
- Published by Elsevier in Research in Autism
- Vol. 132
Exploring trajectories of anxiety symptoms in autistic and non-autistic youth over adolescence using parent- and self-report measures
- 23 January 2026
- Published by Elsevier in Research in Autism
- Vol. 132
Open AccessExploring trajectories of anxiety symptoms in autistic and non-autistic youth over adolescence using parent- and self-report measures
- Open Access
Improved Estimation of Correlation Accuracy for Machine Learning Brain-Phenotype Associations
Machine learning is used in neuroscience to examine brain-phenotype associations and facilitate individual prediction from high-dimensional brain imaging. For continuous phenotypes, Pearson correlation between the observed and predicted phenotype is used to quantify model accuracy in testing data. However, recent research suggests millions of samples may be needed to reliably estimate the maximum achievable predictive accuracy (MAPA). We formally define the MAPA and show that the Pearson estimator is biased for this quantity and its confidence intervals fail to capture the target. We develop a semiparametric (double machine learning) one-step estimator that more accurately estimates the MAPA and yields valid confidence intervals across flexible machine learning settings. Analyzing data from the Reproducible Brain Charts dataset, we show that this estimator has smaller bias when estimating brain-phenotype associations of neuroimaging data with age and psychopathology phenotypes. We show that MAPA for psychopathology factor scores using machine learning models built on structural and functional imaging measures is not better than using demographic and nuisance covariates alone.
Open AccessImproved Estimation of Correlation Accuracy for Machine Learning Brain-Phenotype Associations
Machine learning is used in neuroscience to examine brain-phenotype associations and facilitate individual prediction from high-dimensional brain imaging. For continuous phenotypes, Pearson correlation between the observed and predicted phenotype is used to quantify model accuracy in testing data. However, recent research suggests millions of samples may be needed to reliably estimate the maximum achievable predictive accuracy (MAPA). We formally define the MAPA and show that the Pearson estimator is biased for this quantity and its confidence intervals fail to capture the target. We develop a semiparametric (double machine learning) one-step estimator that more accurately estimates the MAPA and yields valid confidence intervals across flexible machine learning settings. Analyzing data from the Reproducible Brain Charts dataset, we show that this estimator has smaller bias when estimating brain-phenotype associations of neuroimaging data with age and psychopathology phenotypes. We show that MAPA for psychopathology factor scores using machine learning models built on structural and functional imaging measures is not better than using demographic and nuisance covariates alone.
- Open Access
- 31 October 2025
- Published by MIT Press in Imaging Neuroscience
- Vol. 3
Semiparametric confidence sets for cross-sectional and longitudinal neuroimaging
The majority of neuroimaging inference focuses on hypothesis testing rather than effect estimation. With concerns about replicability, there is growing interest in reporting standardized effect sizes from neuroimaging group-level analyses. Confidence sets for effect sizes were recently developed for neuroimaging, but are restricted to simple univariate contrasts (e.g., one-sample or two-sample test Cohen’s d) and cross-sectional data. Thus, existing methods exclude increasingly common longitudinal associations of biological brain measurements with, potentially nonlinear, inter- and intra-individual variations in diagnosis, development, or symptoms. We use modern methods for confidence sets combined with a recently proposed robust effect size index to provide a very general approach and software for effect size confidence set inference in neuroimaging. Our method involves robust estimation of the effect size image and spatial and temporal covariance function based on generalized estimating equations. We use a nonparametric bootstrap to estimate the joint distribution of the robust effect size image across voxels to construct confidence sets. These confidence sets identify regions of the image where the lower or upper simultaneous confidence interval is above or below a given threshold with high probability. We evaluate the coverage and simultaneous confidence interval width of the proposed procedures using realistic simulations and perform longitudinal analyses of aging and diagnostic differences of cortical thickness in Alzheimer’s disease and diagnostic differences of resting-state hippocampal activity in psychosis. This comprehensive approach, along with the visualization functions integrated into the pbj R package, offers a robust tool for analyzing repeated neuroimaging measurements.
- 31 October 2025
- Published by MIT Press in Imaging Neuroscience
- Vol. 3
Open AccessSemiparametric confidence sets for cross-sectional and longitudinal neuroimaging
The majority of neuroimaging inference focuses on hypothesis testing rather than effect estimation. With concerns about replicability, there is growing interest in reporting standardized effect sizes from neuroimaging group-level analyses. Confidence sets for effect sizes were recently developed for neuroimaging, but are restricted to simple univariate contrasts (e.g., one-sample or two-sample test Cohen’s d) and cross-sectional data. Thus, existing methods exclude increasingly common longitudinal associations of biological brain measurements with, potentially nonlinear, inter- and intra-individual variations in diagnosis, development, or symptoms. We use modern methods for confidence sets combined with a recently proposed robust effect size index to provide a very general approach and software for effect size confidence set inference in neuroimaging. Our method involves robust estimation of the effect size image and spatial and temporal covariance function based on generalized estimating equations. We use a nonparametric bootstrap to estimate the joint distribution of the robust effect size image across voxels to construct confidence sets. These confidence sets identify regions of the image where the lower or upper simultaneous confidence interval is above or below a given threshold with high probability. We evaluate the coverage and simultaneous confidence interval width of the proposed procedures using realistic simulations and perform longitudinal analyses of aging and diagnostic differences of cortical thickness in Alzheimer’s disease and diagnostic differences of resting-state hippocampal activity in psychosis. This comprehensive approach, along with the visualization functions integrated into the pbj R package, offers a robust tool for analyzing repeated neuroimaging measurements.
- Open Access
Semiparametric Confidence Sets for Arbitrary Effect Sizes in Longitudinal Neuroimaging
The majority of neuroimaging inference focuses on hypothesis testing rather than effect estimation. With concerns about replicability, there is growing interest in reporting standardized effect sizes from neuroimaging group-level analyses. Confidence sets are a recently developed approach to perform inference for effect sizes in neuroimaging but are restricted to univariate effect sizes and cross-sectional data. Thus, existing methods exclude increasingly common multigroup or nonlinear longitudinal associations of biological brain measurements with inter- and intra-individual variations in diagnosis, development, or symptoms. We broadly generalize the confidence set approach by developing a method for arbitrary effect sizes in longitudinal studies. Our method involves robust estimation of the effect size image and spatial and temporal covariance function based on generalized estimating equations. We obtain more efficient effect size estimates by concurrently estimating the exchangeable working covariance and using a nonparametric bootstrap to determine the joint distribution of effect size across voxels used to construct confidence sets. These confidence sets identify regions of the image where the lower or upper simultaneous confidence interval is above or below a given threshold with high probability. We evaluate the coverage and simultaneous confidence interval width of the proposed procedures using realistic simulations and perform longitudinal analyses of aging and diagnostic differences of cortical thickness in Alzheimer’s disease and diagnostic differences of resting-state hippocampal activity in psychosis. This comprehensive approach along with the visualization functions integrated into the pbj R package offers a robust tool for analyzing repeated neuroimaging measurements.
Open AccessSemiparametric Confidence Sets for Arbitrary Effect Sizes in Longitudinal Neuroimaging
The majority of neuroimaging inference focuses on hypothesis testing rather than effect estimation. With concerns about replicability, there is growing interest in reporting standardized effect sizes from neuroimaging group-level analyses. Confidence sets are a recently developed approach to perform inference for effect sizes in neuroimaging but are restricted to univariate effect sizes and cross-sectional data. Thus, existing methods exclude increasingly common multigroup or nonlinear longitudinal associations of biological brain measurements with inter- and intra-individual variations in diagnosis, development, or symptoms. We broadly generalize the confidence set approach by developing a method for arbitrary effect sizes in longitudinal studies. Our method involves robust estimation of the effect size image and spatial and temporal covariance function based on generalized estimating equations. We obtain more efficient effect size estimates by concurrently estimating the exchangeable working covariance and using a nonparametric bootstrap to determine the joint distribution of effect size across voxels used to construct confidence sets. These confidence sets identify regions of the image where the lower or upper simultaneous confidence interval is above or below a given threshold with high probability. We evaluate the coverage and simultaneous confidence interval width of the proposed procedures using realistic simulations and perform longitudinal analyses of aging and diagnostic differences of cortical thickness in Alzheimer’s disease and diagnostic differences of resting-state hippocampal activity in psychosis. This comprehensive approach along with the visualization functions integrated into the pbj R package offers a robust tool for analyzing repeated neuroimaging measurements.
- 11 February 2025
- Published by SAGE Publications in Statistical Methods in Medical Research
- Vol. 34 (5) , 986-1000
Statistical considerations for evaluating treatment effect under various non-proportional hazard scenarios
We conducted a systematic comparison of statistical methods used for the analysis of time-to-event outcomes under various proportional and non-proportional hazard (NPH) scenarios. Our study used data from recently published oncology trials to compare the Log-rank test, still by far the most widely used option, against some available alternatives, including the MaxCombo test, the Restricted Mean Survival Time difference test, the Generalized Gamma model and the Generalized F model. Power, type I error rate, and time-dependent bias with respect to the survival probability and median survival time were used to evaluate and compare the performance of these methods. In addition to the real data, we simulated three hypothetical scenarios with crossing hazards chosen so that the early and late effects “cancel out” and used them to evaluate the ability of the aforementioned methods to detect time-specific and overall treatment effects. We implemented novel metrics for assessing the time-dependent bias in treatment effect estimates to provide a more comprehensive evaluation in NPH scenarios. Recommendations under each NPH scenario are provided by examining the type I error rate, power, and time-dependent bias associated with each statistical approach.
- 11 February 2025
- Published by SAGE Publications in Statistical Methods in Medical Research
- Vol. 34 (5) , 986-1000
Statistical considerations for evaluating treatment effect under various non-proportional hazard scenarios
We conducted a systematic comparison of statistical methods used for the analysis of time-to-event outcomes under various proportional and non-proportional hazard (NPH) scenarios. Our study used data from recently published oncology trials to compare the Log-rank test, still by far the most widely used option, against some available alternatives, including the MaxCombo test, the Restricted Mean Survival Time difference test, the Generalized Gamma model and the Generalized F model. Power, type I error rate, and time-dependent bias with respect to the survival probability and median survival time were used to evaluate and compare the performance of these methods. In addition to the real data, we simulated three hypothetical scenarios with crossing hazards chosen so that the early and late effects “cancel out” and used them to evaluate the ability of the aforementioned methods to detect time-specific and overall treatment effects. We implemented novel metrics for assessing the time-dependent bias in treatment effect estimates to provide a more comprehensive evaluation in NPH scenarios. Recommendations under each NPH scenario are provided by examining the type I error rate, power, and time-dependent bias associated with each statistical approach.
- 25 January 2024
- Published by Springer Nature in Journal of Systems Science and Complexity
- Vol. 37 (2) , 863-885
Partial Linear Model Averaging Prediction for Longitudinal Data
- 25 January 2024
- Published by Springer Nature in Journal of Systems Science and Complexity
- Vol. 37 (2) , 863-885
Partial Linear Model Averaging Prediction for Longitudinal Data
- Open Access
- 1 April 2022
- Published by IOP Publishing in Journal of Physics: Conference Series
- Vol. 2219 (1) , 012062
Research on Automatic Control Model of Two-wheeled Differential Robot
Aiming at the problem that the speed control of the traditional two-wheeled differential robot is cumbersome and requires a lot of manpower, the intelligent two-wheeled speed matching and intelligent path finding model of the robot under the known surrounding environment are studied. The extended method is used to process the environment image to solve the volume problem of the robot itself, and the improved PRM algorithm is used to solve the shortest path, and the post-alignment of the shortest path is distinguished. Using the radius of curvature, the speed ratio of the two-wheeled differential robot at each position can be obtained, thus establishing the automatic control of the two-wheeled differential robot. The automatic control model of two-wheeled robot is synthesized. The model can effectively reduce the manpower consumed by the robot in path finding and control, and it has important case reference significance for saving the economic cost of two-wheeled differential robot control.
- 1 April 2022
- Published by IOP Publishing in Journal of Physics: Conference Series
- Vol. 2219 (1) , 012062
Open AccessResearch on Automatic Control Model of Two-wheeled Differential Robot
Aiming at the problem that the speed control of the traditional two-wheeled differential robot is cumbersome and requires a lot of manpower, the intelligent two-wheeled speed matching and intelligent path finding model of the robot under the known surrounding environment are studied. The extended method is used to process the environment image to solve the volume problem of the robot itself, and the improved PRM algorithm is used to solve the shortest path, and the post-alignment of the shortest path is distinguished. Using the radius of curvature, the speed ratio of the two-wheeled differential robot at each position can be obtained, thus establishing the automatic control of the two-wheeled differential robot. The automatic control model of two-wheeled robot is synthesized. The model can effectively reduce the manpower consumed by the robot in path finding and control, and it has important case reference significance for saving the economic cost of two-wheeled differential robot control.
- Open Access
- 2 March 2017
- Published by American Institute of Aeronautics and Astronautics (AIAA)
The Phenomena of Sudden Change in Hypersonic Inlets Self-Starting Capability
- 2 March 2017
- Published by American Institute of Aeronautics and Astronautics (AIAA)
Open AccessThe Phenomena of Sudden Change in Hypersonic Inlets Self-Starting Capability