DeepLabCut

…it develops open-source tools such as DeepLabCut and trains biomechanically realistic embodied agents. Position 1 (computer vision and AI for behaviour analysis) covers behavioural datasets for animal groups, animal and environment reconstruction, behaviour underst

…upon two open-source ML platforms, DeepLabCut (Mathis et al. 2018) and Variational Animal Motion Encoding (VAME) (Luxem et al. 2022), to track and analyze naturalistic behavior in App knock-in (KI) and 5xFAD transgenic mouse models of AD. New human

…training machine learning models (DeepLabCut, CEBRA etc.), writing post-hoc analysis code, and/or running visualization tools. Those pre-defined sub-tasks require domain expertise in coding, machine learning, and often heavy manual annotation. Yet, there has not been one system that can help abstract away all the sub-tasks and focus on interaction-based programming to meet researchers’ high-level language queries. Here, inspired by the design of Software-3.0, we introduce a

Animals use their whole body to navigate their environment and solve tasks. Zebrafish (Dario rerio) need to coordinate tail and fin movements and can serve as a model organism for the study of locomotion and locomotion generation and dysregulation. Behavioral tracking of zebrafish has usually restricted itself to tracking in the (x,y) plane with an overhead camera. This precludes studies that investigate postural control or navigation in naturalistic, three-dimensional conditions. To study full body coordination we built a rig that allows us to track larval zebrafish in 3D with high resolution

Alzheimer's disease (AD) is a neurodegenerative disorder that is characterised by the gradual deterioration of executive functions and memory. Its primary pathophysiological hallmarks are amyloid plaques and neurofibrillary tangles. Promising humanised knock-in AD mouse models have been developed that exhibit Aβ pathology, neuroinflammation, synaptic dysfunction, and behavioural impairment. A rigorous battery of behavioural tests and analyses was utilised to further our knowledge of Alzheimer's disease.Our study comprised young (6-months), middle-aged (12-months), and old (22-months) AppNL-G-F

Animals and humans adapt to environmental changes by encoding prior experiences. While classical conditioning paradigms like fear conditioning have been the main focus in the learning and memory field, other forms of associative learning are gaining attention. One example is higher-order conditioning, mediated learning, which allows non-reinforced neutral stimuli to induce attraction or aversion through associations with directly reinforced cues. In rodents, sensory preconditioning, and second-order conditioning tasks have been used to assess higher-order conditioning. In this study, we develo

…We used DeepLabCut to analyze the behaviorof twelve homing pigeons in a 4sqm arena with two separate food patches. We manipulatedthe elevation of the food patches to investigate foraging strategies condi

In our previous study, we conducted a rescue behavior experiment. In this experiment, mice were trained to rescue a distressed familiar conspecific of the same sex by opening a door under two conditions: when the trapped mouse was in a dry or flooded chamber. We demonstrated that mice exhibit rescue behavior towards a distressed conspecific. Additionally, we discovered that inhibiting the dorsal hippocampus impairs this behavior. However, several questions regarding the specific components of impaired behavior remain. Does the emotional arousal of potential helpers increase when witnessing a d

…memory task where mice, tracked with DeepLabCut automated

…markerless pose estimation using DeepLabCut based on deep neural network was applied to ICH model rats that can be selectively inhibited in the regulatory system of motor function.To inhibit the regulatory system after functional recovery, the cerebello-rubral tract is selectively blocked with DREADD

Focused ultrasound (FUS) is a novel noninvasive modality for neuromodulation in the central nervous system with a high spatial resolution and the ability to target deep brain structures. Rodent studies have shown motor responses following FUS but mostly failed to demonstrate consistent and target-specific effects. Here, we used a high-frequency (4 MHz) transducer with a small focal area to investigate the target-specificity of FUS neuromodulation. Mice were implanted with a large cranial window consisting of an ultrasound-transparent Polymethylpentene membrane. For the experiment, awake mice w

Integrating sensory information, such as visual inputs, is critical for many goal-directed behaviors, including motor learning. Motor adaptation, a specific type of adaptive learning where one needs to rapidly learn a new sensorimotor mapping has been an excellent way to behaviorally test how motor learning may be enacted. For example, in humans, perturbing visual feedback results in deviated trajectories that participants can restore to baseline performance over time. Yet, the neural circuits that underlie these visuomotor adaptations remain poorly understood. Here, we developed an implicit v

Sleep quantity and quality change dramatically across the lifespan of altricial mammals born with immature brains such as humans and rodents (Logan 2019, Cirelli 2020, Blumberg 2005, Blumberg 2014). Despite a crucial role for healthy development, we still know quite little about developmental sleep – especially when it comes to architecture and oscillatory dynamics, or to the emergence of global inter-regional brain cross-communication. Current methods to analyse immature sleep fail to provide the resolution needed to characterise its longitudinal microstructure. We designed light-weighted chr

Goal-directed saccades have been studied extensively in primates. In contrast, eye movements in mice have been thought to mainly coordinate eye-head coupling, rather than actively select visual input, due to the absence of a retinal fovea and their reliance on non-visual senses. But does this hold true during naturalistic behaviour?To study mouse eye movements in a naturalistic context featuring self-initiated visual flow and the need for visual target tracking, we designed a foraging task presented in a virtual reality, projected onto a 210-degree immersive environment (Fig1A). The saccades w

Behavioral science, a multidisciplinary field spanning neuroscience, psychology, and ecology, aims to understand individual, collective, and social behaviors along with their neural and genetic correlates. Nevertheless, quantifying naturalistic behaviors remains challenging due to time-consuming manual methods that are prone to errors and inconsistencies. Integrating techniques from physics, computer science, and mathematics has improved this process, particularly in pose estimation. Several tools for pose estimation exist, but they have been mainly designed for standard animal models such as

…Thus, tools like DeepLabCut have built in methods for ReID (Lauer et al. Nature Methods 2022). For non-patterned species, such as mice, gorillas and fruit flies, ReID is particularly challenging due to the lack of salient body markings across individuals. Here we focus on one of the most difficult species: Brown bears (Ursus arctos). ReIDing this species is partic

Social interaction is a core aspect of mammalian behavior, and alterations in social behaviors are found across many neurodevelopmental and psychiatric conditions. Mice display a range of socioemotional behaviors and are commonly used as models to investigate the neuronal circuits and molecular mechanisms underlying differences in social behaviors. Analyzing social interactions in mice is often done by manual quantification of videos. Although multiple methods for automated tracking have been developed, reliable tracking and automated behavioral classification of multiple freely-moving unmarke

Animals’ ability to execute and adapt movements relies on neural networks in the spinal cord that regulate the rhythm and pattern of muscle contractions. Neuronal dysfunctions within these circuits, resulting from physiological or pathological processes, affect postural stability and resilience of motor behavior to external perturbations. To assess the biomechanical strategies adopted by mice to execute postural-demanding motor tasks, we reconstructed 2D kinematics in mice challenged to run on a treadmill and cross increasingly narrower beams. The joint coordinates in markerless mice were trac

…we used the machine learning platforms DeepLabCut and VAME to test the hypoth

Undirected behavior reflects cognitive functions and provides insights for diagnosing psychiatric conditions such as bipolar disorder (BD) (McReynolds, 1962). Open-field animal behaviors have been well-studied for this purpose; however, a corresponding human subject paradigm is still lacking, and quantifying complex spontaneous human behaviors is challenging. Here, we demonstrate a semi-supervised model to quantify undirected human behavior, differentiate subtle hallmark behavioral features of BD, and create a natural language generative model to provide nuanced interpretations of behaviors wi

…Accessible software tools, such as DeepLabCut, have made it possible to include behavioral pose as one data modality in larger neuroscience projects. Staying at the leading edge of neuroscience requires significant data engineering and computational programming skills to handle day-to-day data operations in order to integrate across multiple modalities, each with separate analysis packages. Together, these increase the data management and synchronization bu

…The recent growth of Deeplabcut, an open source toolbox for marker-less pose estimation based on transfer learning with deep neural networks, promises to greatly enhance our ability to track animal pose with high-accuracy and high-efficiency, because it can match human labelling accuracy with minimal training data. Here, we implement Deeplabcut routines and develop a pipeline to accurately track the pose of mice undergoing defensive behaviors. Defensive behaviors are elicited

…kinematics using a single camera and DeepLabCut (or similar) software. A lightweight, transparent polycarbonate floor and a mirror mounted on the inside allow two views to be captured simultaneously. All parts are commercially available or laser cut

Adaptive motor control requires the integration of proprioceptive and other sensory signals. However, the principles, which govern the processing of proprioception, are poorly understood. Here, we employ a task-driven modeling approach to quantitatively test hypotheses about the functional role of proprioceptive neurons in cuneate nucleus, which is the projection target of ascending spindle neurons in the brain stem. We generated datasets of realistic proprioceptive, muscle spindle inputs for a large, diverse repertoire of movements (following Sandbrink et al., bioRxiv, 2020), and used them to