There are many active research projects accessing and applying shared ADNI data. Use the search above to find specific research focuses on the active ADNI investigations. This information is requested annually as a requirement for data access.
| Principal Investigator | |
| Principal Investigator's Name: | Devin Setiawan |
| Institution: | University of Kansas |
| Department: | Engineering |
| Country: | |
| Proposed Analysis: | Early and accurate prediction of AD is crucial for timely intervention and improved patient care. My proposed analysis aims to leverage the rich and comprehensive Alzheimer's Disease Neuroimaging Initiative (ADNI) dataset to develop an interpretable machine learning model for predicting the onset of AD and gain insights into the underlying mechanisms that contribute to the disease's progression. I would use the dataset in a few different ways. First, the dataset will be used to develop a classification model that is highly interpretable. I believe an interpretable model has the advantage of being able to provide an easy-to-understand model that can help further understanding for experts in comparison to black-box models. Secondly, the dataset will be used for feature importance analysis. By analyzing the importance of each feature, I aim to shed light on the underlying biological, genetic, and neuroimaging factors associated with AD, providing valuable insights into the disease's etiology. Lastly, I would also like to explore the interaction effects between different features. The aim is to identify synergistic or antagonistic relationships between various components, contributing to a deeper understanding of the disease mechanisms and potentially revealing novel targets for therapeutic interventions. To accomplish this, I would use state-of-the-art interpretable machine learning models that can produce sparse additive linear models. In terms of feature importance analysis, I would like to bring in some of the new ways to analyze feature importance such as using Shapley values. As for interaction effects, I would like to explore some new statistical methods that had been used in other domains. |
| Additional Investigators |
