2016-08-012024-05-14https://scholars.lib.ntu.edu.tw/handle/123456789/660887摘要：初級視覺皮層V1神經的主要功能之一是偵測空間中的明暗對比(luminance contrast)。明暗對比訊 息由視網膜的節細胞(retinal ganglion cells)負責偵測，然後經由侧膝核(lateral geniculate nucleus)將訊號 傳到V1。視網膜上有兩種不同功能的節細胞：ON-center cells負責偵測正向對比(positive contrast,例如： 白字灰底），OFF-center cells負責偵測負向對比(negative contrast,例如:黑字灰底），這兩個平行的頻道會 在V1聚合，大部分V1神經會同時接收這兩種神經來的訊號。雖然ON-center跟OFF-center cells的 反應強度及數目在視網膜及侧膝核並無明顯差異，許多獮猴V1神經卻顯現明顯的暗偏好：對負向對 比的反應大於正向對比，而且暗偏好在許多動物的V1 (包括貓及人類等)都非常顯著。本研究的主要目標在瞭解V1明暗偏好的神經機制及其對視知覺的影響。目標1 :在獼猴V1產生 暗偏好之神經機制。在V1訊號輸入皮質層(layer-4C)，暗偏好神經在數目上比明偏好神經多，但其產 生的神經機制卻不清楚。我們會使用多頻道電極來同時記錄不同V1皮質層中神經元訊號的明暗偏好 程度，我們會針對4C皮質層中的magnocellular neurons (負責偵測明暗)及parvocellular neurons(負責偵 測色彩），以及其下游皮質層的神經做分析，並使用不同的視覺刺激來研究明暗偏好的情境效果 (contextual effect)。目標2 :暗偏好與神經共振的關聯。當暗偏好訊號在4C皮質層開始減弱，其在V1 輸出皮質層(layer-2/3)卻維持一段時間後才下降，而且強度被強烈放大。一個可能原因是2/3皮質層中 的復發網絡(recurrent network)增強並延長暗偏好的訊號。藉由分析神經間訊號的相關（spike correlation)，我們假設暗偏好神經的共振強度在2/3皮質層會比在4C皮質層強，而且也會大於明偏好 神經間的共振強度。目標3:神經暗偏好如何影響獼猴視知覺。人類對負向對比的敏感度遠大於正向 對比。雖然獮猴V1有較多的暗偏好神經，其強度卻會隨視覺刺激的特性而改變，其對視知覺的影響 也不清楚。我們將訓練清醒且具行為能力的獮猴去偵測視覺刺激，藉由調整刺激亮度來產生不同對比 值(Weber contrast)，測量獮猴對明暗對比的敏感度及偏好，是否會因為刺激特性(例如:靜止或移動)或 微電刺激干擾(microstimulation)而有所不同。目標4 :老鼠明暗偏好及其對視知覺的影響。夜行性的老 鼠其V1視覺神經在功能上與許多日行性的靈長類相似，但其V1卻沒有柱狀組織(columnar organization) 來呈現導向(orientation)及方向性(direction)等靈長類V1所擁有的特徵。最近研究更發現老鼠V1呈現 明顯的明偏好(Polack & Contreras 2012)!為瞭解V1神經偏好對視知覺的影響，我們訓練老鼠去偵測明 暗視覺刺激，並預期老鼠對正向對比的敏感度會大於負向對比。總體而言，這些研究結果有助於瞭解V1如何處裡明暗對比訊息，並對視知覺與V1神經的連結 提供重要線索。動物初級視覺皮層V1的研究，不僅為瞭解人類早期視覺系統提供很好的實驗模型， 同時也對腦皮質層組織在處理訊息的一般原則提供寳貴洞見，研究結果也有助於V1電腦神經網絡及 視覺輔具之建立。<br> Abstract: Detecting contrast is one of the main features for neurons in the primary visual cortex (V1). Contrast is first encoded by retinal ganglion cells, and is sent to V1 via the lateral geniculate nucleus (LGN) of the thalamus. There are mainly two types of retinol ganglion cells: ON-center cells are responsible for detecting positive contrast (white on gray background), and OFF-center cells for detecting negative contrast (black on gray background). The two parallel channels converge in V1, and most V1 neurons receive input from both types of neurons. ON-center and OFF-center cells are about the same in response amplitudes and numbers of cells in the retina and the LGN, but many neurons in macaque V1 have a black-over-white bias: responses to negative contrasts are larger to positive contrasts.The main goal of the study is to understanding the neural mechanism underlying the black bias in V1 and how the bias may affect visual perception. Aim 1: the origin of the black bias in macaque V1. There are more black-dominated than white-dominated neurons in the thalamocortical input layer 4C of macaque V1, but how the black-over-white bias is formed remains unclear. We focus on the difference between magnocellular neurons (detecting luminance) and parvocellular neurons (detecting color), and use different stimulus ensembles to study the contextual effect of the black bias. Aim 2: neural synchrony and the black bias. When the black-dominated signal in layer-4C starts to decline, that in the corticocortical output layers 2/3 is greatly amplified and sustains for a period of time. The recurrent excitation in layers 2/3 may be responsible for amplifying and prolonging the black-dominated signal. We propose to study the spike correlation in different layers of macaque V1 and hypothesize that synchrony among black-dominated neurons is stronger in layers 2/3 than in layer 4C. Aim 3: how V1 black bias affects perception in macaque. Humans are more sensitive to black than to white, but it remains unclear how the black dominance in V1 affects monkey’s perception. Awake and behaving monkeys are trained to detect stimuli with different Weber contrasts. We study how contrast sensitivity in monkeys is affected by the types of stimuli (e.g. static or moving) and by interference with microstimulation. Aim 4: contrast preference and perception in rodents. Rodent V1 is endowed with most major functional cell types identified in higher mammals, but it does not have columnar organization to represent features such as orientation and direction. A recent study even reported that rodent V1 showed greater responses to bright than to dark stimuli (Polack & Contreras 2012)! We will train rats to detect stimuli with different contrast polarities and hypothesize that they prefer positive contrasts to negative contrasts.Overall, these results will provide information about how contrast is processed in V1 and how perceptual bias in contrast is achieved. Animal V1 research not only provides a good experimental model for understanding human early visual system, but also offers invaluable insight about the general rules of cortical organization.主要視覺皮質區(V1)明暗對比神經共振視覺受域神經衝動觸發平均獮猴V1spatial contrastsynchronyreceptive fieldspike-trigger averagemacaque