In this study, we use Deep Learning-based (DL) method to denoising ASL CBF images. Convolutional neural networks with a “wide” structure, residual learning and batch normalization are utilized as the core of our denoising model. Comparing to non-DL-based methods, the proposed method showed a significant SNR increase as well as partial volume effects improvement. Also, the DL-based method requires less CBF input images, which significantly shorten the acquisition time and reduce the chance of head motion.
This abstract and the presentation materials are available to members only; a login is required.