Sebastian Domsch1, Jascha Zapp2, Lothar R Schad2, Frauke Nees3, Holger Hill4, Derik Hermann4, Karl Mann4, Sabine Vollstädt-Klein4. 1. Department of Computer Assisted Clinical Medicine, Medical Faculty Mannheim, Heidelberg University, Germany. Electronic address: sebastian.domsch@medma.uni-heidelberg.de. 2. Department of Computer Assisted Clinical Medicine, Medical Faculty Mannheim, Heidelberg University, Germany. 3. Department of Cognitive and Clinical Neuroscience, Central Institute of Mental Health, Medical Faculty Mannheim, Heidelberg University, Germany. 4. Department of Addictive Behavior and Addiction Medicine, Central Institute of Mental Health, Medical Faculty Mannheim, Heidelberg University, Germany.
Abstract
BACKGROUND: To translate highly accelerated EPI-fMRI protocols as commonly used at ultra-high field strengths to clinical 3T settings. NEW METHOD: EPI protocols with increasing matrix sizes and parallel imaging (PI) factors were tested in two separate fMRI studies, a simple motor-task and a complex motivation-task experiment with focus on the sensorimotor cortex (SMC) and the nucleus accumbens (NAcc), respectively. RESULTS: By increasing the matrix size and the PI-factor simultaneously, BOLD-sensitivity in terms of maximal t-values and numbers of activated clusters was uncompromised in single individuals in both fMRI experiments. In the SMC, the multi-subject analysis revealed an increase of 66% of the maximal t-value whereby the number of activated clusters was increased by a factor of 3.3 when the matrix size (PI-factor) was increased from 96×96 (R=2) to 192×192 (R=4). In the NAcc, the number of activated clusters increased from 5 to 7 whereby the maximal t-value remained unaffected when the matrix size (PI-factor) was increased from 96×96 (R=2) to 160×160 (R=3). COMPARISON WITH EXISTING METHOD: Using the proposed high-resolution EPI protocol, spatial blurring was clearly reduced. Further, BOLD sensitivity was clearly improved in multi-subject analyses and remained unaffected in single individuals compared to using the standard protocols. CONCLUSIONS: Conventionally used matrix sizes (PI-factors) might be non-optimal for some applications sacrificing BOLD spatial specificity. We recommend using the proposed high-resolution protocols applicable in detecting robust BOLD activation in fMRI.
BACKGROUND: To translate highly accelerated EPI-fMRI protocols as commonly used at ultra-high field strengths to clinical 3T settings. NEW METHOD: EPI protocols with increasing matrix sizes and parallel imaging (PI) factors were tested in two separate fMRI studies, a simple motor-task and a complex motivation-task experiment with focus on the sensorimotor cortex (SMC) and the nucleus accumbens (NAcc), respectively. RESULTS: By increasing the matrix size and the PI-factor simultaneously, BOLD-sensitivity in terms of maximal t-values and numbers of activated clusters was uncompromised in single individuals in both fMRI experiments. In the SMC, the multi-subject analysis revealed an increase of 66% of the maximal t-value whereby the number of activated clusters was increased by a factor of 3.3 when the matrix size (PI-factor) was increased from 96×96 (R=2) to 192×192 (R=4). In the NAcc, the number of activated clusters increased from 5 to 7 whereby the maximal t-value remained unaffected when the matrix size (PI-factor) was increased from 96×96 (R=2) to 160×160 (R=3). COMPARISON WITH EXISTING METHOD: Using the proposed high-resolution EPI protocol, spatial blurring was clearly reduced. Further, BOLD sensitivity was clearly improved in multi-subject analyses and remained unaffected in single individuals compared to using the standard protocols. CONCLUSIONS: Conventionally used matrix sizes (PI-factors) might be non-optimal for some applications sacrificing BOLD spatial specificity. We recommend using the proposed high-resolution protocols applicable in detecting robust BOLD activation in fMRI.