The emergence and influence of internal states

Neuron, Год журнала: 2022, Номер 110(16), С. 2545 - 2570

Опубликована: Май 27, 2022

Язык: Английский

---

DeepEthogram, a machine learning pipeline for supervised behavior classification from raw pixels

eLife, Год журнала: 2021, Номер 10

Опубликована: Сен. 2, 2021

Videos of animal behavior are used to quantify researcher-defined behaviors interest study neural function, gene mutations, and pharmacological therapies. Behaviors often scored manually, which is time-consuming, limited few behaviors, variable across researchers. We created DeepEthogram: software that uses supervised machine learning convert raw video pixels into an ethogram, the present in each frame. DeepEthogram designed be general-purpose applicable species, video-recording hardware. It convolutional networks compute motion, extract features from motion images, classify behaviors. classified with above 90% accuracy on single frames videos mice flies, matching expert-level human performance. accurately predicts rare requires little training data, generalizes subjects. A graphical interface allows beginning-to-end analysis without end-user programming. DeepEthogram's rapid, automatic, reproducible labeling may accelerate enhance analysis. Code available at: https://github.com/jbohnslav/deepethogram.

Язык: Английский

---

Natural behavior is the language of the brain

Current Biology, Год журнала: 2022, Номер 32(10), С. R482 - R493

Опубликована: Май 1, 2022

Язык: Английский

---

Identifying behavioral structure from deep variational embeddings of animal motion

Communications Biology, Год журнала: 2022, Номер 5(1)

Опубликована: Ноя. 18, 2022

Abstract Quantification and detection of the hierarchical organization behavior is a major challenge in neuroscience. Recent advances markerless pose estimation enable visualization high-dimensional spatiotemporal behavioral dynamics animal motion. However, robust reliable technical approaches are needed to uncover underlying structure these data segment into discrete hierarchically organized motifs. Here, we present an unsupervised probabilistic deep learning framework that identifies from variational embeddings motion (VAME). By using mouse model beta amyloidosis as use case, show VAME not only motifs, but also captures representation motif’s usage. The approach allows for grouping motifs communities differences community-specific motif usage individual cohorts were undetectable by human visual observation. Thus, segmentation applicable wide range experimental setups, models conditions without requiring supervised or a-priori interference.

Язык: Английский

---

Behavioural and dopaminergic signatures of resilience

Nature, Год журнала: 2022, Номер 611(7934), С. 124 - 132

Опубликована: Окт. 19, 2022

Язык: Английский

---

A paradigm shift in translational psychiatry through rodent neuroethology

Molecular Psychiatry, Год журнала: 2023, Номер 28(3), С. 993 - 1003

Опубликована: Янв. 12, 2023

Abstract Mental disorders are a significant cause of disability worldwide. They profoundly affect individuals’ well-being and impose substantial financial burden on societies governments. However, despite decades extensive research, the effectiveness current therapeutics for mental is often not satisfactory or well tolerated by patient. Moreover, most novel therapeutic candidates fail in clinical testing during expensive phases (II III), which results withdrawal pharma companies from investing field. It also brings into question using animal models preclinical studies to discover new agents predict their potential treating illnesses humans. Here, we focus rodents as propose that they essential investigations candidate agents’ mechanisms action safety efficiency. Nevertheless, argue there need paradigm shift methodologies used measure behavior laboratory settings. Specifically, behavioral readouts obtained short, highly controlled tests impoverished environments social contexts proxies complex human might be limited face validity. Conversely, monitored more naturalistic over long periods display ethologically relevant behaviors reflect evolutionarily conserved endophenotypes translational value. We present how semi-natural setups groups mice individually tagged, video recorded continuously can attainable affordable. open-source machine-learning techniques pose estimation enable continuous automatic tracking individual body parts periods. The trajectories each further subjected supervised machine learning algorithms detection specific (e.g., chasing, biting, fleeing) unsupervised motifs stereotypical movements harder name label manually). Compared animals wild, compatible with neural genetic manipulation techniques. As such, study neurobiological underlying behavior. Hence, suggest such possesses best out classical ethology reductive behaviorist approach may provide breakthrough discovering efficient therapies illnesses.

Язык: Английский

---

Deep-learning-based identification, tracking, pose estimation and behaviour classification of interacting primates and mice in complex environments

Nature Machine Intelligence, Год журнала: 2022, Номер 4(4), С. 331 - 340

Опубликована: Апрель 21, 2022

Язык: Английский

---

Simple Behavioral Analysis (SimBA) as a platform for explainable machine learning in behavioral neuroscience

Nature Neuroscience, Год журнала: 2024, Номер 27(7), С. 1411 - 1424

Опубликована: Май 22, 2024

Язык: Английский

---

Keypoint-MoSeq: parsing behavior by linking point tracking to pose dynamics

bioRxiv (Cold Spring Harbor Laboratory), Год журнала: 2023, Номер unknown

Опубликована: Март 17, 2023

Abstract Keypoint tracking algorithms have revolutionized the analysis of animal behavior, enabling investigators to flexibly quantify behavioral dynamics from conventional video recordings obtained in a wide variety settings. However, it remains unclear how parse continuous keypoint data into modules out which behavior is organized. This challenge particularly acute because susceptible high frequency jitter that clustering can mistake for transitions between modules. Here we present keypoint-MoSeq, machine learning-based platform identifying (“syllables”) without human supervision. Keypoint-MoSeq uses generative model distinguish noise effectively identify syllables whose boundaries correspond natural sub-second discontinuities inherent mouse behavior. outperforms commonly used alternative methods at these transitions, capturing correlations neural activity and classifying either solitary or social behaviors accordance with annotations. therefore renders grammar accessible many researchers who use standard capture

Язык: Английский

---

Keypoint-MoSeq: parsing behavior by linking point tracking to pose dynamics

Nature Methods, Год журнала: 2024, Номер 21(7), С. 1329 - 1339

Опубликована: Июль 1, 2024

Abstract Keypoint tracking algorithms can flexibly quantify animal movement from videos obtained in a wide variety of settings. However, it remains unclear how to parse continuous keypoint data into discrete actions. This challenge is particularly acute because are susceptible high-frequency jitter that clustering mistake for transitions between Here we present keypoint-MoSeq, machine learning-based platform identifying behavioral modules (‘syllables’) without human supervision. Keypoint-MoSeq uses generative model distinguish noise behavior, enabling identify syllables whose boundaries correspond natural sub-second discontinuities pose dynamics. outperforms commonly used alternative methods at these transitions, capturing correlations neural activity and behavior classifying either solitary or social behaviors accordance with annotations. also works multiple species generalizes beyond the syllable timescale, fast sniff-aligned movements mice spectrum oscillatory fruit flies. Keypoint-MoSeq, therefore, renders accessible modular structure through standard video recordings.

Язык: Английский

---