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: | Liang Wenwen |
| Institution: | Huazhong University of Science and Technology |
| Department: | Tongji Hospital, Tongji Medical College |
| Country: | |
| Proposed Analysis: | The model was evaluated using data from the Alzheimer's Disease Neuroimaging Initiative 1 (ADNI-1), and external validation of the model was performed on ADNI-2 to demonstrate the usefulness of the model in the study. A new method using MFPC method and Cox regression model is adopted. The ADNI-1 dataset was first used to investigate whether multivariable longitudinal markers could enhance the predictive performance of AD conversion in distinguishing between high-risk and low-risk patients compared to baseline markers. Two pre-planned Cox regression models were specified, the first model (Model 1) consisting only of baseline observations of five markers, and the second model (model 2) by including longitudinal markers from the interest. In addition to the five baseline neurocognitive markers in model 1, the predictive power of baseline imaging markers was studied by including hippocampal volume, middle temporal lobe volume, and FDG-PET for a comprehensive comparison. Call this model Model 1A. Time-dependent subject operating Characteristic curves (iAUC; To evaluate the performance of each prognostic model. The combined Brier score [22] (the lower the better) was also used as a calibration tool to assess the consistency between predicted and real risks. In order to avoid overestimating predictive performance, the survival model was cross-validated by a factor of 10 (CV). In the second analysis, the structural form of Model 2 was used to assess changes in model performance because of the addition of additional longitudinal markers. In external validation, the parameters estimated from ADNI-1 are used to calculate PI ADNI in ADNI-2. The risk group for conversion from MCI to AD consisted of the quartiles of PI ADNI calculated in ADNI-1. In addition, how to update PI ADNI of hypothetical patients over time as new clinical information becomes available is demonstrated. |
| Additional Investigators |
