A Comprehensive Clinical fMRI Software Solution to Enable Mapping of Critical Functional Networks and Cerebrovascular Reactivity in the Brain

A Comprehensive Clinical fMRI Software Solution to Enable Mapping of Critical Functional Networks and Cerebrovascular Reactivity in the Brain

Our preliminary work demonstrates that an integrated fMRI software solution, incorporating tb-fMRI, rs-fMRI, and CVR mapping, is clinically feasible and helps clinicians plan brain tumor resection. We have developed a novel automated seed selection method that can accurately map language networks from rs-fMRI. We hypothesize that our innovative approach to enhance, optimize, and validate our preliminary software and integrate it with an established fMRI platform will create robust solutions for clinical RSN and CVR mapping. Partnering with NordicNeuroLab (NNL) will leverage the professional software development by a seasoned commercial MRI software producer in coordination with leading clinical and research experts at MD Anderson. The research will be conducted through three specific aims:

1. Develop a clinical software platform for mapping RSNs and determine optimized workflow for presurgical localization of eloquent areas.
2. Develop a clinical software platform for mapping CVR and determine optimized workflow for identifying and visualizing brain areas with potential false-negative fMRI results.
3. Test and validate RSN and CVR mapping software in patients undergoing neurosurgery.

Primary Objectives:

To validate resting-state fMRI (rs-fMRI) mapping of language and motor (hand and tongue) areas obtained by using the software developed in this research: We will test the non-inferiority of rs-fMRI comparing to task-based fMRI (tb-fMRI) by using intra-operative direct cortical stimulation as the gold standard.

To test whether CVR mapping using the software developed in this research will increase confidence of fMRI mapping of language and motor (hand and tongue) areas: We will test whether the accuracy of fMRI, assessed by using intra-operative direct cortical stimulation as the gold standard, will improve after incorporating the CVR mapping.

Secondary Objectives:

To optimize the rs-fMRI and CVR mapping software during the development phases and to test feasibility of the post-processing workflow using clinical images.

Inclusion criteria:

1. patient must be >/= 18 years of age;
2. patients who will undergo neurosurgical resection of brain tumors;
3. patients who will undergo presurgical fMRI as standard of care;
4. patients who on neurosurgical evaluation will undergo direct cortical stimulation (DCS) as standard of care.

For the retrospective study, we will review up to 1600 patients all clinical presurgical fMRI studies performed on brain tumor patients at the MD Anderson since 5/2004.

The inclusion criteria:

1. patients must be >/= 18 years of age;
2. patients who had the presurgical fMRI that contained at least one task-based fMRI scan.

Exclusion Criteria:

Exclusion criteria for the prospective study:

1. patients cannot give informed consent;
2. patients cannot undergo MRI examinations;
3. patients who are indicated for speech fMRI only but cannot comply with any of the speech tasks based on the evaluation of a clinical neuropsychologist.

Exclusion criteria for the retrospective study:

1. all fMRI scans were contaminated by head motions (translation > 2 mm or rotation > 2 degrees);
2. only speech fMRI paradigms were performed, and patients were not able to comply with any of the speech tasks based on the evaluation of a clinical neuropsychologist.