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: | Rehman Tariq |
Institution: | University of Calgary |
Department: | Biomedical Engineering |
Country: | |
Proposed Analysis: | Generative Adversarial Networks (GANs) have been previously utilized to generate novel MRI images (T1, T2, and FLAIR) of healthy individuals. There are even algorithms that will translate a T1 image into a T2 image, and vice versa. My objective is to develop a GAN algorithm that can translate any of T1,T2, or FLAIR sequences into any of the remaining two, while also having the capability to generate new sequences of the same type (Ex. using T1-images to generate a new T1 image). Specifically, I will be modifying the existing cycleGAN algorithm for this study. To accomplish this, I will need a large dataset of individuals who have been diagnosed with a specific disease in order to test whether disease characteristics can be conserved by the GAN algorithm. This will allow for the standardization of clinical MRI images while also presenting the ability to greatly reduce scan times by using a corresponding sequence (such as a T1 sequence) to generate the remaining sequences. Furthermore, given that large datasets are highly likely to have missing sequences, this algorithm can be used to generate missing sequences. |
Additional Investigators |