Fahad Almuqhim

Ph.D. Candidate


Lecture Information:
  • October 30, 2023
  • 10:00 AM
  • CASE 349 & Zoom

Speaker Bio

Fahad Almuqhim is currently a Ph.D. candidate at the Knight Foundation School of Computing & Information Sciences (KFSCIS) at Florida International University, where he is conducting research under the supervision of Prof. Fahad Saeed. His research is primarily focused on enhancing the classification of mental disorders through the development of innovative deep learning models, including Sparse Autoencoders, Attention Networks, and Self-Supervised Architectures, particularly for multi-site fMRI data.

Fahad’s academic journey started with a B.Sc. in Computer Science from Imam University in Saudi Arabia. He later pursued his M.Sc. in Computer Science at the Rochester Institute of Technology (RIT). Since 2011, Fahad has been a lecturer in the Department of Computer Science at Imam University in Saudi Arabia. He has published in the journal Frontiers in Computational Neuroscience, and in Frontiers in Neuroinformatics. Most recently, he was granted a US patent (# 11379981) for his work in July 2022.

Abstract

To date, the diagnostic process for individuals with brain disorders is based purely on behavioral descriptions of symptomology observed in various settings by trained psychologists and psychiatrists. Mental disorders affect the structure or function of the brain and can manifest in a range of symptoms and severity, impacting cognitive function, motor skills, emotions, and overall quality of life.

Autism spectrum disorder (ASD) is a mental disorder that has wide range of symptoms or characteristics such as limited communication (including verbal and non-verbal), limited social interaction, and subjects may exhibit repeated or limited interests. Individuals with ASD have numerous challenges in daily life, and often develop comorbidities such as depression, anxiety disorder, or ADHD which may complicate the diagnostic processes. Early diagnosis of ASD can result in early intervention, which can significantly improve the developmental outcomes especially for young children and families facing socioeconomic difficulties.

Biomarkers are quantifiable characteristics related to brain structure or function and may consist of complex patterns associated with mental disorders. Neuroimaging such as Functional Magnetic Resonance Imaging (fMRI) data has provided a steppingstone for biomarker discovery and can capture data about brain anatomy, connectivity, and operation. fMRI data provides high-dimensional features, which can make it challenging to detect subtle ASD biomarker patterns using conventional computational methods.

Our proposal is to design and develop deep learning models that can detect complex biomarkers to distinguish ASD brains from healthy control (HC) brains using fMRI data. Our proposal consists of three parts: 1) Improving ASD diagnosis by designing and developing deep learning classifier that outperforms current clinical diagnostic performance. 2) Improving the identification of biomarkers and reducing the influence of site-specific signatures leading to generalizable methods. 3) Improving ASD diagnosis by designing and developing a self-supervised model that can be trained on limited labeled data.