Xiaoyu (Sherry) Ding, Ph.D.

@article{Ding2015,

title = {Multivariate classification of smokers and nonsmokers using SVM-RFE on structural MRI images.},

author = {Xiaoyu Ding and Yihong Yang and Elliot A Stein and Thomas J Ross},

url = {https://www.ncbi.nlm.nih.gov/pubmed/26497657},

doi = {10.1002/hbm.22956},

issn = {1097-0193 (Electronic); 1065-9471 (Linking)},

year  = {2015},

date = {2015-10-24},

urldate = {2015-10-24},

journal = {Hum Brain Mapp},

volume = {36},

number = {12},

pages = {4869--4879},

address = {Neuroimaging Research Branch, Intramural Research Program, National Institute on Drug Abuse, National Institutes of Health, Baltimore, Maryland.},

abstract = {Voxel-based morphometry (VBM) studies have revealed gray matter alterations in smokers, but this type of analysis has poor predictive value for individual cases, which limits its applicability in clinical diagnoses and treatment. A predictive model would essentially embody a complex biomarker that could be used to evaluate treatment efficacy. In this study, we applied VBM along with a multivariate classification method consisting of a support vector machine with recursive feature elimination to discriminate smokers from nonsmokers using their structural MRI data. Mean gray matter volumes in 1,024 cerebral cortical regions of interest created using a subparcellated version of the Automated Anatomical Labeling template were calculated from 60 smokers and 60 nonsmokers, and served as input features to the classification procedure. The classifier achieved the highest accuracy of 69.6% when taking the 139 highest ranked features via 10-fold cross-validation. Critically, these features were later validated on an independent testing set that consisted of 28 smokers and 28 nonsmokers, yielding a 64.04% accuracy level (binomial P = 0.01). Following classification, exploratory post hoc regression analyses were performed, which revealed that gray matter volumes in the putamen, hippocampus, prefrontal cortex, cingulate cortex, caudate, thalamus, pre-/postcentral gyrus, precuneus, and the parahippocampal gyrus, were inversely related to smoking behavioral characteristics. These results not only indicate that smoking related gray matter alterations can provide predictive power for group membership, but also suggest that machine learning techniques can reveal underlying smoking-related neurobiology.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{Ding2013b,

title = {Cocaine addiction related reproducible brain regions of abnormal default-mode network functional connectivity: a group ICA study with different model orders.},

author = {Xiaoyu Ding and Seong-Whan Lee},

url = {https://www.ncbi.nlm.nih.gov/pubmed/23707901},

doi = {10.1016/j.neulet.2013.05.029},

issn = {1872-7972 (Electronic); 0304-3940 (Linking)},

year  = {2013},

date = {2013-05-22},

journal = {Neurosci Lett},

volume = {548},

pages = {110--114},

address = {Department of Computer Science and Engineering, Korea University, Anam-dong, Seongbuk-ku, Seoul 136-713, Republic of Korea. xyding@image.korea.ac.kr},

abstract = {Model order selection in group independent component analysis (ICA) has a significant effect on the obtained components. This study investigated the reproducible brain regions of abnormal default-mode network (DMN) functional connectivity related with cocaine addiction through different model order settings in group ICA. Resting-state fMRI data from 24 cocaine addicts and 24 healthy controls were temporally concatenated and processed by group ICA using model orders of 10, 20, 30, 40, and 50, respectively. For each model order, the group ICA approach was repeated 100 times using the ICASSO toolbox and after clustering the obtained components, centrotype-based anterior and posterior DMN components were selected for further analysis. Individual DMN components were obtained through back-reconstruction and converted to z-score maps. A whole brain mixed effects factorial ANOVA was performed to explore the differences in resting-state DMN functional connectivity between cocaine addicts and healthy controls. The hippocampus, which showed decreased functional connectivity in cocaine addicts for all the tested model orders, might be considered as a reproducible abnormal region in DMN associated with cocaine addiction. This finding suggests that using group ICA to examine the functional connectivity of the hippocampus in the resting-state DMN may provide an additional insight potentially relevant for cocaine-related diagnoses and treatments.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{Ding2013,

title = {Changes of functional and effective connectivity in smoking replenishment on deprived heavy smokers: a resting-state FMRI study.},

author = {Xiaoyu Ding and Seong-Whan Lee},

url = {https://www.ncbi.nlm.nih.gov/pubmed/23527165},

doi = {10.1371/journal.pone.0059331},

issn = {1932-6203 (Electronic); 1932-6203 (Linking)},

year  = {2013},

date = {2013-03-19},

journal = {PLoS One},

volume = {8},

number = {3},

pages = {e59331},

address = {Department of Computer Science and Engineering, Korea University, Anam-dong, Seongbuk-ku, Seoul, Republic of Korea.},

abstract = {Previous researches have explored the changes of functional connectivity caused by smoking with the aid of fMRI. This study considers not only functional connectivity but also effective connectivity regarding both brain networks and brain regions by using a novel analysis framework that combines independent component analysis (ICA) and Granger causality analysis (GCA). We conducted a resting-state fMRI experiment in which twenty-one heavy smokers were scanned in two sessions of different conditions: smoking abstinence followed by smoking satiety. In our framework, group ICA was firstly adopted to obtain the spatial patterns of the default-mode network (DMN), executive-control network (ECN), and salience network (SN). Their associated time courses were analyzed using GCA, showing that the effective connectivity from SN to DMN was reduced and that from ECN/DMN to SN was enhanced after smoking replenishment. A paired t-test on ICA spatial patterns revealed functional connectivity variation in regions such as the insula, parahippocampus, precuneus, anterior cingulate cortex, supplementary motor area, and ventromedial/dorsolateral prefrontal cortex. These regions were later selected as the regions of interest (ROIs), and their effective connectivity was investigated subsequently using GCA. In smoking abstinence, the insula showed the increased effective connectivity with the other ROIs; while in smoking satiety, the parahippocampus had the enhanced inter-area effective connectivity. These results demonstrate our hypothesis that for deprived heavy smokers, smoking replenishment takes effect on both functional and effective connectivity. Moreover, our analysis framework could be applied in a range of neuroscience studies.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{Ding2012,

title = {Performance evaluation of nonnegative matrix factorization algorithms to estimate task-related neuronal activities from fMRI data.},

author = {Xiaoyu Ding and Jong-Hwan Lee and Seong-Whan Lee},

url = {https://www.ncbi.nlm.nih.gov/pubmed/23200679},

doi = {10.1016/j.mri.2012.10.003},

issn = {1873-5894 (Electronic); 0730-725X (Linking)},

year  = {2012},

date = {2012-11-30},

journal = {Magn Reson Imaging},

volume = {31},

number = {3},

pages = {466--476},

address = {Department of Computer Science and Engineering, Korea University, Anam-dong, Seongbuk-ku, Seoul, Korea. xyding@image.korea.ac.kr},

abstract = {Nonnegative matrix factorization (NMF) is a blind source separation (BSS) algorithm which is based on the distinct constraint of nonnegativity of the estimated parameters as well as on the measured data. In this study, according to the potential feasibility of NMF for fMRI data, the four most popular NMF algorithms, corresponding to the following two types of (1) least-squares based update [i.e., alternating least-squares NMF (ALSNMF) and projected gradient descent NMF] and (2) multiplicative update (i.e., NMF based on Euclidean distance and NMF based on divergence cost function), were investigated by using them to estimate task-related neuronal activities. These algorithms were applied firstly to individual data from a single subject and, subsequently, to group data sets from multiple subjects. On the single-subject level, although all four algorithms detected task-related activation from simulated data, the performance of multiplicative update NMFs was significantly deteriorated when evaluated using visuomotor task fMRI data, for which they failed in estimating any task-related neuronal activities. In group-level analysis on both simulated data and real fMRI data, ALSNMF outperformed the other three algorithms. The presented findings may suggest that ALSNMF appears to be the most promising option among the tested NMF algorithms to extract task-related neuronal activities from fMRI data.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}