No abstract

Transcranial magnetic stimulation (TMS) over occipital cortex can impair visual processing. Such “TMS masking” has repeatedly been shown at several stimulus onset asynchronies (SOAs), with TMS pulses generally applied after the onset of a visual stimulus. Following increased interest in the neuronal state-dependency of visual processing, we recently explored the efficacy of TMS at “negative SOAs”, when no visual processing can yet occur. We could reveal pre-stimulus TMS disruption, with results moreover hinting at two separate mechanisms in occipital cortex biasing subsequent orientation perception. Here we extended this work, including a chronometric design to map the temporal dynamics of spatially specific and unspecific mechanisms of state-dependent visual processing, while moreover controlling for TMS-induced pupil covering. TMS pulses applied 60–40 ms prior to a visual stimulus decreased orientation processing independent of stimulus location, while a local suppressive effect was found for TMS applied 30–10 ms pre-stimulus. These results contribute to our understanding of spatiotemporal mechanisms in occipital cortex underlying the state-dependency of visual processing, providing a basis for future work to link pre-stimulus TMS suppression effects to other known visual biasing mechanisms.

Dendritic spines are the seat of most excitatory synapses in the brain, and a cellular structure considered central to learning, memory, and activity-dependent plasticity. The quantification of dendritic spines from light microscopy data is usually performed by humans in a painstaking and error-prone process. We found that human-to-human variability is substantial (inter-rater reliability 82.2 ± 6.4 %), raising concerns about the reproducibility of experiments and the validity of using human-annotated 'ground truth' as an evaluation method for computational approaches of spine identification. To address this, we present DeepD3, an open deep learning-based framework to robustly quantify dendritic spines in microscopy data in a fully automated fashion. DeepD3's neural networks have been trained on data from different sources and experimental conditions, annotated and segmented by multiple experts and they offer precise quantification of dendrites and dendritic spines. Importantly, these networks were validated in a number of datasets on varying acquisition modalities, species, anatomical locations and fluorescent indicators. The entire DeepD3 open framework, including the fully segmented training data, a benchmark that multiple experts have annotated, and the DeepD3 model zoo is fully available, addressing the lack of openly available datasets of dendritic spines while offering a ready-to-use, flexible, transparent, and reproducible spine quantification method.

Eye-specific segregation of retinal ganglion cell (RGC) axons in the dorsal lateral geniculate nucleus (dLGN) is considered a hallmark of visual system development. However, a recent anatomical study showed that nearly half of the neurons in dLGN of adult mice still receive input from both retinae, but functional data about binocularity in mature dLGN is conflicting. Here, we found that a variable but small fraction of thalamocortical neurons is binocular in vivo. Using dual-channel optogenetics in vitro we correspondingly found that dLGN neurons are dominated by retinogeniculate input from one eye only, although most neurons also received small but detectable input from the non-dominant eye. Anatomical overlap between RGC axons and dLGN neuron dendrites did not explain this strong bias towards monocularity. Our data rather suggest that functional input selection and refinement, leaving the remaining non-dominant eye inputs in a juvenile-like state, underlies the prevalent monocularity of neurons in dLGN.