Graduation Semester and Year
Spring 2024
Language
English
Document Type
Dissertation
Degree Name
Doctor of Philosophy in Computer Science
Department
Computer Science and Engineering
First Advisor
Dajiang Zhu
Second Advisor
Junzhou Huang
Third Advisor
Jean Gao
Fourth Advisor
Li Wang
Abstract
When we review the history of development of artificial intelligence (AI), we will find that brain science plays a pivotal role in fostering breakthroughs in AI, such as artificial neural networks (ANNs). Today, AI has made remarkable strides, particularly with the emergence of large language models (LLMs), surpassing expectations and achieving human-level performance in certain tasks. Nonetheless, an insurmountable gap remains between AI and human intelligence. It is urgent to establish a bridge between brain science and AI, promoting their mutual enhancement and collaborations. This involve establishing connections from brain science to AI (brain-inspired AI), and reversely, from AI to brain (AI in brain science).
For this purpose, in this thesis, we aim to 1) take advantages of the powerful capabilities of cutting-edge AI techniques to explore the intricacies of the brain, including both normal brain and brain disorders; and 2) leverage the superior organizational principles of brain networks to inspire and guide the design of AI models. My research in the two directions has opened new frontiers for brain science and AI research: 1)Exploring the foundational organizational principles of human brain with AI techniques. We develop expressive and effective deep learning models that can capture the non-trivial brain structure–function relationship at individual level; 2) Identifying imaging-based biomarkers of brain dementia and modeling the continuous brain disease progression using AI techniques. We develop multi-modal deep neural networks to integrate multiple types of brain network connectome and characterize their deep relationship as an “individual connectome signature” for brain disease study. We also propose novel structure learning methods to model the continuum of disease progression; 3) LLMs in neuroscience and healthcare domain. LLMs are at the forefront of AI, which have change the paradign of model design. To explore how to unlock the potential benefits of LLMs in neuroscience and improve healthcare outcomes, we have developed effective approaches to tailor these models to the unique requirements of the specialized domains. 4) Pioneering research in brain-inspired AI. We conduct post hoc analysis to explore the connections between artificial neural networks (ANNs) and biological neural networks (BNNs), laying a solid foundation for brain-inspired AI. Building upon these findings, we proactively instill the organizational principles observed in BNNs into ANNs and develop more powerful ANNs.
Keywords
Brain science, Artificial intelligence, Brain inspired AI, AI in neuroscience, Large language models in healthcare
Disciplines
Artificial Intelligence and Robotics | Bioinformatics | Computational Neuroscience | Data Science | Disease Modeling
License
This work is licensed under a Creative Commons Attribution-NonCommercial-No Derivative Works 4.0 International License.
Recommended Citation
Zhang, Lu, "When Brain Meets Artificial Intelligence" (2024). Computer Science and Engineering Dissertations. 2.
https://mavmatrix.uta.edu/cse_dissertations/2
Included in
Artificial Intelligence and Robotics Commons, Bioinformatics Commons, Computational Neuroscience Commons, Data Science Commons, Disease Modeling Commons