Measurement of stretch-evoked brainstem function using fMRI

Abstract

Knowledge on the organization of motor function in the reticulospinal tract (RST) is limited by the lack of methods for measuring RST function in humans. Behavioral studies suggest the involvement of the RST in long latency responses (LLRs). LLRs, elicited by precisely controlled perturbations, can therefore act as a viable paradigm to measure motor-related RST activity using functional Magnetic Resonance Imaging (fMRI). Here we present StretchfMRI, a novel technique developed to study RST function associated with LLRs. StretchfMRI combines robotic perturbations with electromyography and fMRI to simultaneously quantify muscular and neural activity during stretchevoked LLRs without loss of reliability. Using StretchfMRI, we established the muscle-specific organization of LLR activity in the brainstem. The observed organization is partially consistent with animal models, with activity primarily in the ipsilateral medulla for flexors and in the contralateral pons for extensors, but also include other areas, such as the midbrain and bilateral pontomedullary contributions. 2/25reference, the mean timeseries of the EMG recorded during Exp 1 and processed using three different filtering pipelines is shown in Fig. 2A. Reliability of stretch-evoked EMG during fMRIWe validated the stretch-evoked EMG collected during fMRI by quantifying agreement between measurements collected inside and outside the MR scanner (see Online Methods). We used two analyses: one based on the mean LLR responses averaged across multiple repetitions for each set of conditions (group-level analysis), and one based on the analysis of EMG measurements of individual perturbations (perturbation-specific analysis). Group level resultsGroup-level analysis was performed using the Bland-Altman (BA) method 13,14 , whose plots are shown in Fig. 2B for the comparisons OUT 2 vs. OUT 1 and IN 1 vs. OUT 1 , for different signal filtering pipelines. The values of bias, positive and negative limits of agreement (LoA), with the respective confidence intervals are reported in Tab. S8 in the supplementary materials.Based on the BA analysis, Adaptive Noise Cancellation (ANC) enables reliable estimation of LLR amplitude during fMRI sessions. In fact, the agreement between LLR amplitudes measured inside and outside the MRI is not worse than the agreement between two OUT sessions. Specifically, the bias estimated for the IN 1 vs. OUT 1 comparison was not significantly different from the one estimated for the OUT 2 vs. OUT 1 comparison (z(72)=0.56, p = 0.57 for stretch, z(72) = 1.70 , p = 0.09 for shortening). Additionally, the 95% confidence intervals of the bias estimated for both comparisons, in both muscle stimulus directions, intersect the zero value indicating that the measurement is unbiased (Fig. 3A). Finally, the overlap between the range defined by the LoA for two repeated OUT experiments and the range of LoA for IN vs. OUT experiment, quantified by the Jaccard coefficient, approached perfect overlap (mean ± s.e.m. J = 0.843 ± 0.001 for stretch, J = 0.783 ± 0.001...

Sensorimotor feedback loops are selectively sensitive to reward

“…A corollary to that proposal is that feedback loops that do not show increased feedback gains with reward would also more generally not be susceptible to pre-modulation like we observe in the LLR (Beckley et al, 1991; de Graaf et al, 2009; Takei et al, 2021; Zonnino et al, 2021). A task in our study that is suited to test this possibility is the cursor jump task, because it does not show feedback gain modulation, while the target jump task, which has a very similar design, does.…”

“…For instance, the circuitry underlying the short-latency rapid response is entirely contained in the spinal cord (Liddell and Sherrington, 1924). The long-latency rapid response relies on supraspinal regions such as the primary motor and primary sensory cortices (Cheney and Fetz, 1984; Day et al, 1991; Evarts and Tanji, 1976; Palmer and Ashby, 1992; Pruszynski et al, 2011a), and is modulated by upstream associative cortical regions (Beckley et al, 1991; de Graaf et al, 2009; Omrani et al, 2016; Takei et al, 2021; Zonnino et al, 2021). Visuomotor feedback responses rely on visual cortex and other cortical and subcortical brain regions (Day and Brown, 2001; Desmurget et al, 2004; Pruszynski et al, 2010).…”