Neurodegeneration is a progressive loss of neuron function or structure, including death of neurons, and occurs at many different levels of neuronal circuitry. In this thesis I discuss Parkinson’s Disease (PD), the second most common neurodegenerative disease (NDD). PD is a devastating progressive NDD often with delayed diagnosis due to detection methods that depend on the appearance of visible motor symptoms. By the time cardinal symptoms manifest, 60 to 80 percent or more of the dopamine-producing cells in the substantia nigra are irreversibly lost. Although there is currently no cure, earlier detection would be highly beneficial to manage treatment and track disease progression. However, today’s clinical diagnosis methods are limited to subjective evaluations and observation. Onset, symptoms and progression significantly vary from patient to patient across stages and subtypes that exceed the scope of a standardized diagnosis. The goal of this thesis is to provide the basis of a more general approach to study the brain, investigating early detection method for NDD with focus on PD. It details the preliminary development, testing and validation of tools and methods to objectively quantify and extrapolate motor and non-motor features of PD from behavioral and cognitive output during everyday life. Measures of interest are categorized within three domains: the motor system, cognitive function, and brain activity. This thesis describes the initial development of non-intrusive tools and methods to obtain high-resolution movement and speech data from everyday life and feasibility analysis of facial feature extraction and EEG for future integration. I tested and validated a body sensor system and wavelet analysis to measure complex movements and object interaction in everyday living situations. The sensor system was also tested for differentiating between healthy and impaired movements. Engineering and design criteria of the sensor system were tested for usability during everyday life. Cognitive processing was quantified during everyday living tasks with varying loaded conditions to test methods for measuring cognitive function. Everyday speech was analyzed for motor and non-motor correlations related to the severity of the disease. A neural oscillation detection (NOD) algorithm was tested in pain patients and facial expression was analyzed to measure both motor and non-motor aspects of PD. Results showed that the wearable sensor system can measure complex movements during everyday living tasks and demonstrates sensitivity to detect physiological differences between patients and controls. Preliminary engineering design supports clothing integration and development of a smartphone sensor platform for everyday use. Early results from loaded conditions suggest that attentional processing is most affected by cognitive demands and could be developed as a method to detect cognitive decline. Analysis of speech symptoms demonstrates a need to collect higher resolution spontaneous speech from everyday living to measure speech motor and non-motor speech features such as language content. Facial expression classifiers and the NOD algorithm indicated feasibility for future integration with additional validation in PD patients. Thus this thesis describes the initial development of tools and methods towards a more general approach to detecting PD. Measuring speech and movement during everyday life could provide a link between motor and cognitive domains to characterize the earliest detectable features of PD. The approach represents a departure from the current state of detection methods that use single data entities (e.g.one-off imaging procedures), which cannot be easily integrated with other data streams, are time consuming and economically costly. The long-term vision is to develop a non-invasive system to measure and integrate behavioral and cognitive features enabling early detection and progression tracking of degenerative disease.

<p dir="ltr">Multi-omic studies ranging from genomics, transcriptomics (e.g., gene expression) to proteomics data exploration have been widely applied to interpret findings from genome wide association studies (GWAS) of Alzheimer's disease (AD). However, previous studies examine each -omics data type individually and the functional interactions between genetic variations, genes and proteins are only used after discovery to interpret the findings, but not beforehand. In this case, multi-omic findings are likely not functionally related and therefore it is challenging for result interpretation. To handle this challenge, we present new modularity constrained least absolute shrinkage and selection operator (M-LASSO), new modularity constrained logistic regression (M-Logistic), new interpretable multi-omic graph fusion neural network model (MoFNet) and new transfer learning framework integrated graph fusion neural network model (TransFuse) to integrate prior biological knowledge to model the functional interactions of multi-omic data. These approaches aim to identify functional connected sub-networks predictive of AD. In this thesis, the intrepretable model MoFNet and TransFuse incorporate prior biological connected multi-omics network, and for the first time model the dynamic information flow from deoxyribonucleic acid (DNA) to ribonucleic acid (RNA) and proteins. While applying the proposed models on multi-omic data from the religious orders study/memory and aging project (ROS/MAP) cohort, MoFNet and TransFuse outperformed all other state-of-art classifiers. Instead of targeting individual markers, the proposed methods identified multi-omic sub-networks associated with AD. MoFNet and TransFuse, produced sub-network and pathway findings that were robustly validated in another independent cohort. These identified gene/protein networks highlight potential pathways involved in AD pathogenesis and could offer systematic overview for understanding the molecular mechanisms of the disease. Investigating these identified pathways in more detail could help uncover the mechanisms causing synaptic dysfunction in AD and guide future research into potential therapeutic targets.</p>

No description available.