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: | Christos Xanthis |
| Institution: | Lund University |
| Department: | Cardiac MR group |
| Country: | |
| Proposed Analysis: | Artificial Intelligence (AI) techniques have become increasingly successful over the last few years in medical image analysis and radiology. However, these techniques come with a major drawback, which either prevents their further development or delays their application in clinical practice. This drawback is related to the availability, relativeness and size of the training data sets required by the associated learning algorithms. The research community has long recognized the importance of this topic. In the field of Magnetic Resonance Imaging (MRI), the process of creating training datasets is considered costly, both in time and money. Creating training datasets requires not only acquisition of medical images that cover a broad spectrum of medical parameters but also manual annotation of the medical images by experts. In addition, this process requires the availability of an MRI scanner and the availability of personnel (such as technologists and radiologists), whereas the complex nature of the underlying MRI physics hinders this process even more. Furthermore, image acquisition is performed under a specific MRI protocol and MRI system configuration. While supervised learning techniques have demonstrated good performance on relatively controlled experiments with standardized imaging protocols, their performance may deteriorate in cases where images are acquired with a different imaging protocol or on a different MRI system. Finally, acquisition of data that cover the entire range of parameters that describe the anatomy and physiology of the population becomes practically impossible. The focus of my research projects is on the fields of advanced MR simulations and synthetic MR datasets that can be used for training AI algorithms. The primary objective of the proposed project is the utilization of the MR protocol that was used for the acquisition of the ADNI dataset on the entire database of synthetic/digital brain models that we have built and the comparison of the realism of the synthetic MR data against the true MR images from the ADNI dataset. Last, advanced AI algorithms will be trained with both synthetic and true MR data and their performance will be evaluated on tasks, such as automatic segmentation. |
| Additional Investigators |
