Ongoing Investigations

ADNI data is made available to researchers around the world. As such, there are many active research projects accessing and applying the shared ADNI data. To further encourage Alzheimer’s disease research collaboration, and to help prevent duplicate efforts, the list below shows the specific research focus of the active ADNI investigations. This information is requested annually as a requirement for data access.

Principal Investigator  
Principal Investigator's Name: Ivan Arisi
Institution: Fondazione EBRI "Rita Levi-Montalcini"
Department: Genomics
Proposed Analysis: We are a group of scientists from different backgrounds, with a common interest in neurodegeneration: IT scientists, bioinformatics, biologists, neurologists. The expertise of our project partners includes: mathematical modelling, software and hardware design, neuroimaging data processing, bioinformatics of –omics data, biochemistry of body fluids, clinical neurology. We propose to integrate existing knowledge, newly generated data and prediction models. The goal is to develop an innovative data analysis and modelling platform for a better, more reliable and more efficient diagnosis, in the framework of the highly complex clinical heterogeneity of neurodegeneration and co-morbidities. Another key point is the discovery of new biomarkers (as measures of disease process as defined by NIH Biomarkers Definitions Working Group, 2001), based on the statistical analysis and data mining of clinical variables and –omics data in patients affected by neurodegeneration. These biomarkers will allow early diagnosis of specific form of dementia, crucial for the choice of adequate preventive measures and a correct drug therapy since the first stages. The objectives of our project are: 1)Standardization and harmonization of clinical and molecular data collection and protocols: clinical examinations, imaging and blood samples collection. 2)Development and application of systems medicine approach by a data integration computational platform: clinical data, neuroimaging and –omics data will be combined. -Creation of a database for retrospective and newly collected data, providing a user-interface query systems for clinicians and scientists, based on data mining methodologies. -Creation of the interactive system to analyse and support study of neurodegenerative diseases. -Creation of a final diagnosis and therapy prediction system for understanding the complexity of clinical phenotypes and co-morbidities, for refining disease models. 3)Identification of mRNA, and proteins in blood samples of patients or specific clinical parameters as early disease biomarkers. 4)Validation of computational predictions (diagnosis, therapy, prevention criteria) in independent well-phenotype patients’ cohorts, provided by our clinical partners, taking into account gender. We are applying to the three resources ADNI, AIBL and DOD-ADNI, since each of them provides curated datasets that would be ideal to set up and train our models and algorithms: -ADNI provides highly curated neuroimaging and genetics data on neurodegeneration. -AIBL is focused on the connection between lifestyle and neurodegeneration. The lifestyle is an absolute priority in the therapeutic approach for Alzheimer’s Disease, due to the lack of effective drug therapies, therefore we may include a structured lifestyle model in our system. -DOD-ADNI would provide the chance to include specific co-morbidities in our models. We would process the large amount of multi-level and multi-scale data from these repositories with machine learning algorithms, such as Logic Mining (Arisi et al, J Alz Dis 2011) , and other methodologies of recent generation developed our IT colleagues, that will allow an easy management of the knowledge extracted from data in the form of diagnostic models for a disease stratification. We would mine the datasets with algorithms able to combine multivariate data of heterogeneous origin: clinical assessment, neuroimaging, molecular, -omics data. Your datasets would be essential to us to train the learning algorithms, due to the excellence and high standardization of your protocols, that would limit the prediction "noise", that is the statistical unreliability of numerical predictions solutions, which in our experience is the major problem when dealing with multicentric studies.
Additional Investigators