This axis focuses on innovations related to the treatment of pain in patients, and technological innovations developed in the form of tools for preclinical research.

Transcranial magnetic stimulation (rTMS) of the primary motor cortex at high frequency (>5 Hz) induces analgesic effects, probably by activating pain modulation systems. We show that a new rTMS paradigm, using continuous theta stimulation (pcTBS), composed of bursts of three pulses at 50 Hz repeated five times per second
for 1 min and 44 s (1,200 pulses), is significantly more effective than rTMS at 10 Hz (Moisset et al. Brain Stim 2015). The pcTBS protocol could therefore have considerable clinical potential, being 8 times shorter but producing stronger analgesic effects than classic rTMS.

Source : Moisset et al. Eur J Neurol. 2017

Lumbar puncture (LP) has been frequently performed for over a century. This procedure is always stressful and often painful. A randomized controlled trial was conducted in adults who required cerebrospinal fluid analysis to evaluate the effectiveness of a fixed mixture of nitrous oxide and 50% oxygen compared to placebo to reduce immediate pain and anxiety procedures during LP (Moisset et al. Eur J Neurol. 2017). We found that inhalation of a fixed 50% N2 O-O2 mixture is effective in reducing PL-induced pain and anxiety. In addition to innovations for patients, several tools have been developed by the laboratory team, with the aim of automating certain repetitive and time-consuming tasks, and therefore reducing inter and intra operator variations. The major needs mainly concern video ethology techniques, or calcium imaging or confocal imaging. In the case of imaging, this involves detecting and counting the neurons in a slice or section of the brain or spinal cord. These neurons are visible either because they have been marked with a fluorescent antibody or because they have been genetically modified to fluoresce upon absorption of calcium.

Delineation of neurons in calcium imaging and deconvolved neuronal activity signal

In the case of calcium imaging, the tool transforms a dynamic 3D image (2D+time) into temporal neuronal activity. The deconvolution operation uses the FOOPSI approach, a constrained sparse deconvolution.

Neural detection by shape optimization

In the case of confocal imaging and fluorescent labeling, detection uses a parametric ellipsoidal model and unconstrained minimization, with an approach resolutely based on the shape of neurons, rather than on fluorescence intensity.

Finally, with the aim of automating the application of the grimace scale in the rating of spontaneous painful behaviors, the team is investing in the disciplinary field of supervised learning, initially to automatically detect the presence of the face of an animal. Preliminary results show excellent detection rates (>90%).