• Select the area you would like to search.
  • ACTIVE INVESTIGATIONS Search for current projects using the investigator's name, institution, or keywords.
  • EXPERTS KNOWLEDGE BASE Enter keywords to search a list of questions and answers received and processed by the ADNI team.
  • ADNI PDFS Search any ADNI publication pdf by author, keyword, or PMID. Use an asterisk only to view all pdfs.
Principal Investigator  
Principal Investigator's Name: Ales Psaker
Institution: The Governor's School at Innovation Park
Department: Science
Proposed Analysis: I am a teacher at The Governor's School at Innovation Park and George Mason University affiliate, and I am the mentor of three high school seniors. Since they are still in high school and do not have the necessary credentials to request the data, I am filling the form out on their behalf. Here is their current proposal: “We are looking to apply machine learning to early detection of Alzheimer’s with this ADNI database. We currently propose employing a multimodal approach to the early detection of Alzheimer’s, with special consideration for 3D MRI images, PET scans, and gene expression data. This combination would allow for greater accuracy in detection. By combining various data sets, a more comprehensive approach can be taken to improve the F1 score of the learning algorithm. Additionally, we would like to investigate the need and concern of missing values among the data. By imputing missing values, the data set can be more comprehensive and stronger correlations can be formed. In regards to the structure of the machine learning algorithm, we plan on using binary classification with SoftMax through a system of 3D (to analyze the 3D MRI and PET scans) and 2D neural networks. Furthermore, we plan on exploring the utilization of SVMs with different kernels to determine the optimal approach for this application. A combination of neural network and an SVM is hypothesized to result in the greatest accuracy. PCA may be used to change the dimensions of the data based on the size of the data set to decrease computational cost while retaining accuracy.”
Additional Investigators