
Background Conventional rodent models for the study of corneal pain commonly evoke eye blink reflex using methods that indiscriminately activate polymodal nociceptors, mechanoreceptors, and thermoreceptors at temporal resolutions that don’t closely match the sub-second timescale of underlying neural dynamics. New method We introduce a novel automated behavioral paradigm for detecting blink reflexes in transgenic TRPV1-ChR2-EYFP mice, enabled by millisecond-precision optogenetic stimulation of TRPV1-lineage corneal afferents (490 nm light). Multi-feature quantification enables univariate and multivariate classification. Results TRPV1-ChR2-EYFP mice (N = 5) blinked in 86.3 ± 10.9% of trials following corneal stimulation with high-intensity (4.69–4.77 mW/mm²) blue light (490 nm, 10 ms pulses) versus 2.5 ± 2.5% at low intensity (<0.01–0.07 mW/mm²). Red light (638 nm) produced no intensity-dependent change. Wild-type mice (N = 3) exhibited none or minimal blink responses to blue-light stimulation. Noxious air puff evoked blinks in 98.8% of all trials. DeepLabCut-based pose estimation extracted six features quantifying the blink reflex, enabling automated detection using univariate and multivariate classifiers, with performance varying by feature and thresholding method. Multivariate classification achieved 99.2% accuracy under four-fold cross-validation and 98.2% mean accuracy under leave-one-mouse-out cross-validation. Automated thresholds yielded classification accuracy comparable to manually-validated thresholds. Comparison with existing methods Unlike conventional air puff paradigms, this optogenetic approach enables temporally precise stimulation of TRPV1-lineage corneal afferents and automated analysis of blink responses at sub-second resolution. Conclusions This video-based method enables accurate, high-throughput, and observer-independent empirical assessment of stimulus-evoked blink reflexes, providing a quantitative platform for investigating corneal sensory responses relevant to ocular nociception.
The human fetal cortex begins as a smooth spheroid and progressively folds into a stereotypical sulcal pattern. I show that the spherical-harmonic expansion — the closed-surface generalisation of Fourier's series — reproduces this developmental sequence in reverse on the group-average atlas: truncating the fully formed cortex at low maximum harmonic degree L_max recovers a younger fetal brain; progressively higher L_max recovers the primary, secondary and tertiary folds in the order in which they emerge. From this I derive a closed-form 23-dimensional log-fractional spectral descriptor. On atlas templates (17 Harvard CRL + 16 dHCP weekly templates) it predicts gestational age at MAE 0.13 and 0.38 wk across cohorts after a single supervised linear alignment, and 1.07 and 1.51 wk without. On 28 Yale-Wayne State longitudinal pairs it advances predicted GA in every subject, without cohort matching. On 80 FeTA scans (31 neurotypical, 49 pathological) off-trajectory distance discriminates pathology at AUC 0.80, and unsupervised clustering yields three internally stable, biometrically concordant subgroups — an exploratory finding requiring external validation. Individual-scan gestational-age accuracy is not the descriptor's competitive use (MAE 2.25 wk on FeTA-NT); its individual-scan value is the shape of the residual around the normative trajectory. The framework separates two spatial-scale regimes of folding onto distinguishable spectral bands and is complementary to existing measures of fetal cortical maturation.
Background Rett syndrome is a genetic neurodevelopmental disorder caused by mutations of the MECP2 gene. It is characterized by severe neurological, cognitive, autonomic, and motor dysfunctions. Central catecholaminergic deficits, together with alterations in GABAergic and glutamatergic systems, have previously been identified in the brain of Mecp2 knockout mice using conventional data analyses. Here, we reanalyzed neurochemical data from this well-established mouse model using machine-learning approaches to identify new neurochemical information. New method We leveraged a neurochemical database featuring tissue concentrations of dopamine, noradrenaline, serotonin, GABA and glutamate from micro-dissected brain areas in both Mecp2-knockout and wild-type mice. We then extracted qualitative and quantitative data at the early-onset and fully symptomatic stages of the disease. Five machine learning models were trained to predict genotype, and their predictive performance was evaluated. Results Despite the small number of individuals per group, the best model was obtained with XGBoost fitting. SHAP-derived variables of importance analysis identified neurotransmitters and brain regions that were consistent with neurochemical alterations previously implicated in the pathophysiology of Rett syndrome. Comparison with existing methods Our results indicate that machine learning, which remains largely underused in the field of neurochemistry, can effectively predict group differences from brain neurochemical profiles in a well-established mouse model for Rett syndrome. Conclusions The satisfying performances highlight the potential of machine learning and modern computational tools to extract additional information from existing neurochemical datasets, complement conventional analyses, and for the redesign of future neurochemical studies.
Background Quantitative magnetic resonance imaging (qMRI) provides insights into tissue microstructure and physiology, making it a tool in preclinical research. However, its processing often requires coding expertise. New method We present Resomapper, an open-source, cross-platform tool designed to simplify multiparametric MRI analysis. Resomapper integrates libraries into a user-friendly workflow that supports T1, T2, and T2* relaxometry mapping, magnetization transfer imaging, and diffusion tensor imaging. It incorporates preprocessing steps -including denoising, Gibbs artifact removal, and bias correction- and enables interactive and automated processing. Results Resomapper was applied to a multiparametric 7 T dataset of healthy mouse brain, generating T1, T2, T2*, magnetization transfer ratio and diffusion tensor maps. The values fell within the ranges reported in the literature and reproduced the differences between grey and white matter regions, illustrating its use in a preclinical study. Comparison with existing methods Resomapper distinguishes itself from qMRI tools not through novel algorithms, but through its integration of established, validated libraries into a single, streamlined pipeline for data conversion, preprocessing and parametric mapping. Maps showed excellent agreement with qMRLab for T2 and were consistent with DSI Studio for mean and axial diffusivity, while systematic differences in radial diffusivity and fractional anisotropy were attributable to the tensor fitting methods. Its design prioritizes reproducibility and standardization, and its Python implementation makes the tool straightforward to install, inspect, and adapt, even for researchers without extensive coding backgrounds. Conclusions By improving accessibility, reproducibility, and standardization, Resomapper aims to facilitate broader adoption of qMRI in preclinical imaging despite current methodological challenges.