Answers are grounded in this paper and cite their sources.
PURPOSE: To evaluate whether a dense, small-element receive array optimized for visual cortex imaging improves temporal SNR and BOLD sensitivity relative to a whole-head coil at 7 T, and whether this depends on spatial resolution. Two custom-built 7 T coils (32 receive, 8 transmit channels), one whole-head and one visual-cortex-optimized, were compared at 0.8- and 2.0-mm resolution. METHODS: Image SNR, temporal SNR, and BOLD sensitivity (suprathreshold spatial extent) were compared between coils at each resolution, acquired as a partial-slab protocol over the occipital lobe. RESULTS: The visual cortex coil yielded 2.5- and 2.8-fold higher image SNR at 0.8- and 2.0-mm resolution, respectively. At 0.8-mm, this only translated into a 1.3-fold increase in temporal SNR, but produced a 4.8-fold greater spatial extent of suprathreshold activation. Although contrast-to-noise-ratio (CNR) was only 2.4-fold higher for the visual cortex coil, decomposing CNR across the suprathreshold t-value distribution showed a 5.5-fold greater weighted CNR contribution, consistent with the larger spatial extent. At 2.0-mm resolution, temporal SNR was 1.4-fold higher with the visual cortex coil, but CNR and activation extent were similar between coils. CONCLUSION: The visual cortex coil's sensitivity advantage was higher at sub-millimeter resolution, where higher image SNR translated into higher temporal SNR and a 4.8-fold greater BOLD detection sensitivity, consistent with a 5.5-fold greater weighted CNR contribution across the suprathreshold t-value distribution. At 2.0-mm resolution, the SNR advantage did not translate into a detection benefit. The visual coil offers a practical advantage for ultra-high-field fMRI targeting fine-scale functional organization in the human visual cortex.
No takes yet. Share an insight, caveat, or question.
Akbari et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: