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: | sahar Nikbakht Aali |
Institution: | University of california, Irvine (UCI) |
Department: | Electrical Engineering and Computer Science(EECS) |
Country: | |
Proposed Analysis: | I am a Ph.D. student at UCI. My field of study is using deep learning techniques in biomedical images. In our project, we want to use deep neural network architecture to reduce the noise in medical images especially PET images. To acquire high- quality PET images, a standard dose of radioactive tracer should be injected into the patients which will lead to a higher risk of radiation damage. On the other hand, dose reduction will affect PET image quality. To solve this problem, we are trying to use machine learning and deep learning algorithms and architectures to improve the image quality for low- dose PET images. We already defined our architecture which is a model of constitutional auto-encoder to reconstruct and denoise the PET images. We are using TensorFlow and Keras libraries for coding purposes. We have a very small dateset of PET images and by using transfer learning methods, good results have been obtained. We are currently in contact with Dr. David Keator and send him our first results. Hopefully, by receiving more datasets we will be able to improve our model and publish our work. |
Additional Investigators |