Document Type


Date of Award


Degree Name

Doctor of Philosophy in Biomedical Engineering - (Ph.D.)


Biomedical Engineering

First Advisor

Xiaobo Li

Second Advisor

Tara L. Alvarez

Third Advisor

Bharat Biswal

Fourth Advisor

Yiyan Liu

Fifth Advisor

Kevin Pang


Attention-deficit/hyperactivity disorder (ADHD) is one of the most prevalent neurodevelopmental disorders in children. Symptoms of childhood ADHD persist into adulthood in around 65% of patients, which elevates the risk for a number of adverse outcomes, resulting in substantial individual and societal burden. A neurodevelopmental double dissociation model is proposed based on existing studies in which the early onset of childhood ADHD is suggested to associate with dysfunctional subcortical structures that remain static throughout the lifetime; while diminution of symptoms over development could link to optimal development of prefrontal cortex. Current existing studies only assess basic measures including regional brain activation and connectivity, which have limited capacity to characterize the functional brain as a high performance parallel information processing system, the field lacks systems-level investigations of the structural and functional patterns that significantly contribute to the symptom remission and persistence in adults with childhood ADHD. Furthermore, traditional statistical methods estimate group differences only within a voxel or region of interest (ROI) at a time without having the capacity to explore how ROIs interact in linear and/or non-linear ways, as they quickly become overburdened when attempting to combine predictors and their interactions from high-dimensional imaging data set.

This dissertation is the first study to apply ensemble learning techniques (ELT) in multimodal neuroimaging features from a sample of adults with childhood ADHD and controls, who have been clinically followed up since childhood. A total of 36 adult probands who were diagnosed with ADHD combined-type during childhood and 36 matched normal controls (NCs) are involved in this dissertation research. Thirty-six adult probands are further split into 18 remitters (ADHD-R) and 18 persisters (ADHD-P) based on the symptoms in their adulthood from DSM-IV ADHD criteria. Cued attention task-based fMRI, structural MRI, and diffusion tensor imaging data from each individual are analyzed. The high-dimensional neuroimaging features, including pair-wise regional connectivity and global/nodal topological properties of the functional brain network for cue-evoked attention process, regional cortical thickness and surface area, subcortical volume, volume and fractional anisotropy of major white matter fiber tract for each subject are calculated. In addition, all the currently available optimization strategies for ensemble learning techniques (i.e., voting, bagging, boosting and stacking techniques) are tested in a pool of semi-final classification results generated by seven basic classifiers, including K-Nearest Neighbors, support vector machine (SVM), logistic regression, Naïve Bayes, linear discriminant analysis, random forest, and multilayer perceptron.

As hypothesized, results indicate that the features of nodal efficiency in right inferior frontal gyrus, right middle frontal (MFG)-inferior parietal (IPL) functional connectivity, and right amygdala volume significantly contributed to accurate discrimination between ADHD probands and controls; higher nodal efficiency of right MFG greatly contributed to inattentive and hyperactive/impulsive symptom remission, while higher right MFG-IPL functional connectivity strongly linked to symptom persistence in adults with childhood ADHD. The utilization of ELTs indicates that the bagging-based ELT with the base model of SVM achieves the best results, with the most significant improvement of the area under the receiver of operating characteristic curve (0.89 for ADHD probands vs. NCs, and 0.9 for ADHD-P vs. ADHD-R). The outcomes of this dissertation research have considerable value for the development of novel interventions that target mechanisms associated with recovery.



To view the content in your browser, please download Adobe Reader or, alternately,
you may Download the file to your hard drive.

NOTE: The latest versions of Adobe Reader do not support viewing PDF files within Firefox on Mac OS and if you are using a modern (Intel) Mac, there is no official plugin for viewing PDF files within the browser window.