Multiscale Gradient Organization of the Human Pulvinar Complex: Integrating Functional Connectivity, Structural Connectivity, and Receptor Expression
The human pulvinar complex exhibits a continuous, gradient-like organization of functional connectivity, structural connectivity, and receptor expression patterns that aligns with the hierarchical organization of cortico-cortical connectivity.
뇌 영상 데이터의 그래프 제약 분석
다변량 기능 데이터에 대한 그래프 제약 분석 방법을 제안하였다. 이는 기존의 기능 가우시안 그래프 모델(FGGM)이 알려진 그래프 구조를 보존할 수 없는 한계를 극복하기 위한 것이다.
Predictive Coding Enables Biologically Plausible Learning of Grid Cells in the Medial Entorhinal Cortex
Predictive coding, a biologically plausible learning rule, can train neural networks to develop hexagonal grid cell representations from spatial inputs, providing a novel and plausible perspective on the learning mechanisms underlying grid cells.
The Exponential Distance Rule Accurately Predicts the Topology and Functional Properties of the Drosophila Connectome
The exponential distance rule, which states that the probability of axons decreases exponentially with their length, accurately describes the structural brain network of the Drosophila fruit fly. A one-parameter model based on this rule effectively captures numerous binary and weighted properties of the Drosophila connectome, revealing that geometric constraints play a key role in shaping brain networks across species.
Modeling Visual Illusions Induced by Localized Stimuli Using Neural Fields
The paper aims to investigate the theoretical modeling of visual illusions observed in Billock and Tsou's experiments, where a localized funnel pattern stimulus induces an orthogonal tunnel pattern in the surrounding region of the visual field. The authors use an Amari-type neural field equation to model the cortical dynamics and explore the role of excitatory and inhibitory neuronal activities in reproducing these nonlinear visual phenomena.
Closed-Loop Seizure Detection and Suppression via Model-Free Control and Algebraic Differentiation in a Noisy Environment
This work proposes a closed-loop framework for detecting and suppressing epileptiform seizures using model-free control techniques and time-domain signal processing methods, which are robust to noise and do not require precise mathematical modeling.
Likelihood of Dendritic Computations in Randomly Connected Feedforward Neural Networks
Random feedforward connectivity, along with large enough neural ensembles (∼100 neurons for groups and ∼1000 neurons for sequences), is sufficient to lead to the convergence of groups or sequences consisting of 3-5 inputs onto small dendritic zones. This provides a substrate for downstream networks to decode arbitrary input patterns.
Modeling Neural Switching Behavior Using Drift-Diffusion Processes
Individual neurons can switch between encoding different stimuli over time, a phenomenon known as neural multiplexing, which offers a scalable encoding scheme to preserve information from distinct stimuli.
神経回路図の解釈による視覚機能の予測
神経回路図の解析から、ショウジョウバエ視覚系の3つのDm3細胞型と3つのTmY細胞型が形態視に関与する回路の一部であると予測された。さらに、これらの細胞の受容野特性や、ロバストな方位選択性、位置不変性、ノイズや錯視の輪郭の補完などの機能が予測された。
Predicting Visual Function from Neuronal Wiring Diagrams in the Drosophila Optic Lobe
Analyzing the structural connectivity of the Drosophila optic lobe can provide theoretical predictions about its functional capabilities in visual processing.
© 2024 by Linnk AI