• 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: Anil Kemisetti
Institution: Genentech
Department: INTONATE
Proposed Analysis: I am part of projects at Roche which aims to create a sustainable health ecosystem by establishing collaborations with pioneers in Neuroimmunology research, able to advance understanding of multiple sclerosis, and apply the insights to create exponential leaps in MS diagnosis monitoring and treatment. Federated Learning and Federated Analytics(FL/FA) play a pivotal role in establishing these collaborations. At the core, FL/FA reduced the privacy risk enabling collaboration providing accessibility to data without incurring the costs involved in creating a centralized database. These privacy guarantees are inherent to FL/FA because it offers a decentralized form of model training or data aggregation, which does not require the data to leave the site. Instead, only the model parameters or the aggregates leave the site. The project is currently working on MVP to establish the Federated Learning Platform. Its primary goal is to evaluate a FA/FL vendor using public datasets followed by Roche/Site Data. The project is looking for a public dataset to conduct a feasibility study involving descriptive statistics, statistical analysis, machine learning, and deep learning on the vendor platform. This study aims to compare the results of the centralized analyses with that of the results obtained from similar analyses using FA/FL. Our team has chosen the ADNI dataset to be one of the public datasets. We plan to use the ADNI dataset and share the assessment results with the ADNI community. Even though the dataset is not related to the MS disease area, we hope our assessment will showcase FL/FA techniques on a vendor platform and inspire new research ideas in Alzheimer’s Disease. Further, in our report, we could present privacy risk mitigation guarantees that FL/FA offers to the community, allay the fears, and encourage the patient community to share the data. I request your permission to use this dataset for our assessment of the vendor platform.
Additional Investigators