Description:

Abstract:

MR-based Neuroimaging has a number of signal “species” that have been developing over the years. These methods are constantly improving from hatchlings, to fingerlings, to trophies due to new hardware, software, pulse sequences, and better analysis pipelines. The “go-to” methods are still providing plenty of fight for neuroimaging projects. Neuroimaging applications and study designs are becoming more complex requiring a better understanding of data reliability for short term and longitudinal studies. The impact of different scanners, scanner type, and different protocols will be discussed along with a possible new tool to mitigate differences. Another topic to be discussed is the “honey hole” known as resting state. There are concerns about different sources of noise and how it might influence the interpretation of connectivity data. Clearly, gross motion such a wiggling, jigging or deep dives can influence the MR signal at that moment and for volumes that follow. How you censor, scrub or scale your data is important for generating results. More subtle variations in the data can result from heart rate and respiration. I will discuss how these confounds can alias into acquired data and review possible actions to mitigate their effects. Next, machine learning applications to neuroimaging data will be discussed. Machine learning is going to revolutionize imaging like the plastic worm did to bass fishing – a game changer! Finally, I will switch reels and discuss a new method that may improve neurosurgical outcomes as well as shed some “light” on BOLD signal behavior. Keeping your license current on neuroimaging methods can prevent you from fishing for results in the depths of your data.

Speaker Info:

Dr. Parrish has a MS/BS in Biomedical Engineering from Case Western Reserve University and a PhD in Biomedical Engineering and Biophysics from the University of Minnesota focused on MR imaging including acquisition, reconstruction, and analysis methods. At the University of Minnesota, he was involved in some early developments of functional MRI. He was later recruited to Northwestern University in 1995 to develop an MR Neuroimaging research program. Since that time, he has become the Director of the Neuroimaging facility that houses 2 research dedicated magnets. He has been successful in obtaining and administering his own research projects investigating: the impact of cerebrovascular disease on brain activation studies, the physiologic basis of the blood oxygen level dependent signal using caffeine and other vaso-challenges, how physiologic noise affects BOLD resting state networks, neuroplasticity changes induced by therapy in chronic stroke, simultaneous collection of perfusion and permeability information in brain tumors, the early detection of Alzheimer’s through changes in the blood brain barrier, and the application of machine learning to brain imaging data for segmentation, classification, and prediction of recovery. Dr. Parrish is interested in all things MRI, especially as they apply to neuroimaging. His recent interest includes the use of infrared imaging for real-time detection of brain function during neurosurgical procedures. He is an avid fisherman!