Surgical resection of brain neoplasms aims at maximizing long-term survival by gross total resection of tumor tissue, while preserving the patient's functional status. Task-based functional MRI (fMRI) is increasingly used as an adjunct to intra-operative electrocorticography for mapping eloquent cortex, but suffers a significant failure rate due to functional impairment or limited task-compliance of patients, and due to nonspecific co-activation. Resting state fMRI (rsfMRI) is a rapidly expanding task-free approach for presurgical mapping, which can efficiently map a large number of resting state networks (RSNs) in clinically feasible scan times. However, clinical acceptance of rsfMRI is impeded by the high degree of expertise required analyzing the data, the high sensitivity to movement, which obscures networks as well as creates false-positive connections, and by temporal and spatial non-stationary of RSNs. The majority of presurgical mapping studies employ seed-based connectivity (SBC) due to its high sensitivity and straightforward interpretation of connectivity. However, variability inherent in investigator-specific and subject-specific seed selection has limited reliability. Thus, there is an urgent need for fully automated whole brain SBC mapping tools with improved tolerance to movement and confounding signal changes in addition to unbiased subject-specific seed selection methods. In the clinical setting this will increase acceptance and reduce cost. The objectives of this phase I STTR research effort are (a) to develop a fully automated real-time resting state fMRI analysis tool that enables online monitoring of data quality and computation of patient specific maps of resting state connectivity, and (b) assess the performance of this tool for robust presurgical mapping in patients with brain tumors. Method development will be based on our sliding window correlation with cumulative meta-statistics approach towards SBC that is highly tolerant to movement and confounding signal sources. In preliminary studies, using this approach in combination with high-speed resting-state fMRI we have demonstrated advantages of mapping eloquent cortex using rsfMRI compared with task-based fMRi in patients with brain lesions. The long-term goal of this research is to improve the clinical utility of resting state fMRI for presurgical mapping by developing an analysis tool that can be seamlessly integrated into the workflow of presurgical planning. This methodology will have significant clinical and commercial potential for a wide range of neurological and psychiatric applications beyond pre-surgical mapping of brain tumors, and will open up examination of brain functional networks to patient populations that have been difficult or impossible to study in the past, such as children and Parkinson's patients.
Imaging of functional connectivity in the human brain using resting state functional MRI (rsfMRI) is a rapidly expanding approach for efficiently mapping multiple brain systems in clinically feasible scan times that does not require patient participatio. This approach is expected to have significant clinical impact for mapping brain function in patients with brain tumors to guide surgical resection. There is now an urgent need for automated tools that can reliably map resting state connectivity in single patients. In the clinica setting this will increase acceptance and reduce cost. The main objective of this STTR research effort is to improve the clinical utility of resting state fMRI for presurgical mapping by developig a fully automated real-time resting state fMRI analysis tool that enables online monitoring of data quality and reliable computation of patient specific maps of resting state connectivity, and to demonstrate feasibility of presurgical mapping in patients with brain tumors using this tool. This methodology will have significant clinical and commercial potential for a wide range of neurological and psychiatric applications beyond pre-surgical mapping of brain tumors, and will open up examination of brain functional networks to patient populations that have been difficult or impossible to study in the past, such as children and Parkinson's patients.
