Chapter 2 Introduction to the Physiology of Perception.

Presentation on theme: "Chapter 2 Introduction to the Physiology of Perception."— Presentation transcript:

1 Chapter 2 Introduction to the Physiology of Perception

2 The Brain: The Mind’s Computer
■Neuron Theory (Golgi,1873) (also see Fig. 2.2)

3 The Brain: Basic Structure
■Doctrine of specific nerve energies (Mueller ,1842) ■Modular organization coordinating information received from two or more senses

4 Neurons: Structure & Electrical Signals
■Five kinds of the receptor neurons: specialized neurons that respond to specific kinds of energy

5 Neurons: Action Potentials
◆Recording electrode is inside the nerve fiber. Reference electrode is outside the fiber. ◆Difference in charge between inside and outside fiber is -70 mV (the resting potential).

6 Neurons: Action Potentials

7 Neurons: Action Potentials - - Two Kinds of Neurotransmitters

8 Neurons: Action Potentials - - Two Kinds of Neurotransmitters
stimulus strength↑  the rate↑ & the regularity of nerve firing↑

9 神經迴路（neural circuits）能登錄複雜的視覺訊息
可用以登錄長度(大小)訊息的聚斂迴路(convergence circuit） Figure Left: A linear circuit with no convergence. Right: Response of neuron B as we increase the number of receptors stimulated.

10 Figure 3. 17 Circuit with convergence added
Figure 3.17 Circuit with convergence added. Neuron B now receives inputs form all of the receptors, so increasing the size of the stimulus increases the size of neuron B’s response.

11 Figure 3. 18 Circuit with convergence and inhibition
Figure 3.18 Circuit with convergence and inhibition. Because stimulation of the receptors on the side (1, 2, 6, and 7) sends inhibition to neuron B, neuron B responds best when just the center (3 - 5) are stimulated.

12 接受域（receptive field, RF）
受刺激時影響神經元激發速率的網膜區域 測試視神經的一個神經元的接受域特性 單細胞記錄（single cell recording）

13 接受域（receptive field, RF）
為何研究接受域? 協助我們瞭解在視覺（系統不同區域的）神經元會受到何種環境刺激的影響，i.e.,會處理何種環境刺激的相關訊息？

14 節細胞接受域形狀為中央-周圍（Center-surround）構成的環狀
閃光刺激A區域或處於黑暗→無反應，resting activation 閃光刺激B區域（興奮區）→ 激發速率變快(on response) 閃光刺激C區域（抑制區）→ 激發速率變慢 但關掉光刺激時，激發速率驟增（off response） --中央為產生on-response的區域，周邊環形為產生off-response的區域，此種接受域稱為on-centered RF or excitatory-center-inhibitory-surround RF 反之，off-centered RF or inhibitory-center-excitatory-center RF What is “dark” for the ganglion cell? 對特定型態的光刺激產生反應→filter（濾器） 100萬個節細胞的接受域佔據視野不同空間位置

15

16 中央周邊的拮抗(antagonism) 中央與周邊相反的效應會彼此抵銷 大小剛好落入中央興奮區的刺激有最佳反應—可用來偵測亮度的差異

17 刺激的表徵 – 「祖母細胞」（grandmother cell）？
感官碼（sensory code） 代表知覺經驗的神經訊號模式 專一登錄（specificity coding） 由只對該刺激作反應的神經訊號代表該刺激 祖母細胞 問題 需要神經元的數量太多 實質上發現一個神經元往往對於多種（類似）刺激有反應

18 Figure 4.26 How faces could be coded according to (a) specificity coding and (b) distributed coding. The height of the bars indicates the response of neurons 1, 2, and 3 to each stimulus face. See text for explanation.

19 分散登錄（distributed coding）
由一組神經元的激發組型來代表每個個別的刺激 優點 少數神經元（彼此合作）就可以代表大量刺激 證據 同一物體引發猴子IT各區域的激發 同類物體引發不同人類腦區域的激發（但仍具模組的概念） 物體的特性除了識別外，還有如顏色，運動，深度等，這些由其他腦區域負責，但最終各層面訊息還是會結合而形成統整的知覺經驗

20 Figure 4.27: Areas in monkey’s IT cortex, determined by a brain scanning procedure, that were activated by Sesame Street’s Ernie (red), a doll’s face (yellow), and a teddy bear (blue). Fig. 4-27, p. 88

21 Figure 4.28a: fMRI responses of the human brain to various types of stimuli. (a) Areas that were most strongly activated by houses, faces, and chairs. Fig. 4-28a, p. 88

22 Figure 4.28b: fMRI responses of the human brain to various types of stimuli. (b) All areas activated by each type of stimulus. Fig. 4-28b, p. 88

23 新近證據—grandmother cell?
Quiroga et al. (2005)運用因癲癇而埋入電極的8位病人的資料 發現有只會對Jennifer Aniston照片反應的單一神經元，也有只對Halle Berry或雪梨歌劇院有反應的單一神經元 但是它們落於海馬迴 - 與記憶經驗有關？

24

25 Sensory Code: Representation of Environment
■Grandmother cell: responding to a specific stimulus, such as a movie star, a concept…etc.. ◆Problems with specificity coding because (1) too many different stimuli to assign specific neurons and (2) most neurons respond to a number of different stimuli. ◆Distributed coding: a particular object represented by the firing of groups of neurons. ◆How many neurons are needed for an object in distributed coding? ▲Sparse coding suggests that only a relatively small number of neurons are necessary to respond to a specific object. This kind of coding can be viewed as a midpoint between specificity and distributed coding.

26 Mind-body Problem How do physiological processes become transformed into perceptual experience? Easy problem of consciousness Neural correlate of consciousness (NCC): How physiological responses correlate with experience ? Hard problem of consciousness How do physiological responses cause experience ?

27 End

28 網膜上神經元匯聚的狀況 網膜上有五種神經元 垂直連結
桿狀、錐狀細胞→兩極細胞(bipolar cell)→神經節細胞(ganglion cell) 水平連結 水平細胞(horizontal cell) 無軸突細胞(amacrine cell)

29 Figure 3.1(a)

30 光線先經過其他網膜的神經元，才抵達桿狀及錐狀細胞
怎麼看得清楚呢？ 透明組織 中央錐細胞密度高處較薄

31 雖然受器細胞有一億二千六百萬個，神經節細胞只有一百萬個，convergence ratio=126:1
Fovea的錐細胞多半與神經節細胞有1-1連結提高錐細胞敏銳度

32 側抑制（lateral inhibition）
用Limulus作實驗動物 每個小眼連結一個受器（刺激一個小眼引發一個受器反應） 鄰近受器的激發產生抑制作用

33 側抑制影響亮度知覺—Hermann Grid

34 Figure 3.6, page 50 Hermann grid
Copyright © 2002 Wadsworth Group. Wadsworth is an imprint of the Wadsworth Group, a division of Thomson Learning

35 與A連結的兩極細胞受到較多側抑制 與B連結的兩極細胞受到較少側抑制
Figure 3.7, page 51 Hermann grid可以用側抑制解釋 與A連結的兩極細胞受到較多側抑制 Copyright © 2002 Wadsworth Group. Wadsworth is an imprint of the Wadsworth Group, a division of Thomson Learning 與B連結的兩極細胞受到較少側抑制

36 側抑制影響亮度知覺—Mach Bands Figure 3.8, page 52 Match bands

37 Figure Circuit to explain the Mach band effect based on lateral inhibition. The circuit works like the one for the Hermann grid in Figure 3.6, with each bipolar cell sending inhibition to its neighbors. If we know the initial output of each receptor and the amount of lateral inhibition, we can calculate the final output of the receptors. (See text for a description of the calculation.)

38 Figure A plot showing the final receptor output calculated for the circuit of figure The bump at B and the dip at C correspond to the light and dark Mach bands, respectively.

39 側抑制影響亮度知覺 -- simultaneous contrast

40 Figure 3.13 How lateral inhibition has been used to explain the simultaneous contrast effect.

41 Figure 3.14 The Benary cross.
這些無法由側抑制解釋 Figure White’s illusion. (From Perception, 1981, 10, p , fig 1a, p Reprinted with permission from Pion, Ltd., London.)

42 神經迴路（neural circuits）能登錄複雜的視覺訊息
可用以登錄長度(大小)訊息的聚斂迴路(convergence circuit） Figure Left: A linear circuit with no convergence. Right: Response of neuron B as we increase the number of receptors stimulated.

43 Figure 3. 17 Circuit with convergence added
Figure 3.17 Circuit with convergence added. Neuron B now receives inputs form all of the receptors, so increasing the size of the stimulus increases the size of neuron B’s response.

44 Figure 3. 18 Circuit with convergence and inhibition
Figure 3.18 Circuit with convergence and inhibition. Because stimulation of the receptors on the side (1, 2, 6, and 7) sends inhibition to neuron B, neuron B responds best when just the center (3 - 5) are stimulated.

45 網膜→初級視皮質（V1, area 17, striate cortex）
Crossover - 10% of visual input goes to - eye movement

46

47 visual pathway left VF of left eye  nasal fibers  cross at optic chiasm  right hemisphere right VF of left eye  temporal fibers  doesn’t cross at optic chiasm  left hemisphere left VF of right eye  temporal fibers  doesn’t cross at optic chiasm  right hemisphere right VF of right eye  nasal fibers  cross at optic chiasm  left hemisphere

48 Crossed 與uncrossed fiber的比例隨眼睛的位置而異
前方：1:1（人） Superior Colliculus （20%） visual center for lower species（eg.,青蛙，魚） 處理where not capable of detailed analysis Orienting, eye movements, spatial localization, visual reflexes

49 側膝核（Lateral Geniculate Nucleus, LGN)
Thalamus 的一部分，層狀結構 RF is center-surround (like ganglion cells) regulates the flow of information in the cortex Receives more information back from the cortex than it receives from the retina Transmit only 4/10 of the nerve impulse from the retina Organizes information

50 Organization in LGN Inputs from two eyes are segregated ipsilateral fibers from ganglion cells  layer 2, 3, 5 contralateral fibers from ganglion cells layer 1,4,6

51 Organization in LGN Retinotopic map 每層之內
各層之間 retintopic map in each layer of LGN are aligned

52 Organization in LGN Different types of LGN cells are segregated layer 1, 2: magnocellular division (large cells) Large cells, respond to brief burst of firing, movement layer 3-6: parvocellular division small cells, respond to sustained st., sustained firing Color, texture/pattern, depth Schiller’s monkey

53 Retinal ganglion cell could also be categorized into three groups (see table) (M-magnocellular, P-parvocellular, K-kinocellular) K cells are particularly small and are responsive to colors (like P cells) M, P & K cells constitute over 90% of the total ganglion cells – cortical areas others – superior colliculus

54

55 The striate cortex primary visual cortex, Area 17, V1
25,000,000 cells (1,000,000 cells in LGN) Organized into layers

56 接受域型態與Hubel & Wiesel（1981諾貝爾醫學獎得主）的預期不同

57

58 Figure 3. 23 (a) The receptive field of a simple cortical cell
Figure 3.23 (a) The receptive field of a simple cortical cell. This cell responds best to a vertical bar of light that covers the excitatory area of the receptive field. The response decreases as the bar is tilted so that it also covers the inhibitory area. (b) Orientation tuning curve of a simple cortical cell for a neuron that responds best to a vertical bar (orientation = 0). (From “Receptive Fields of Single Neurons in the Cat’s Striate Cortex, “ by D. H. Hubel and T. N. Wiesel, 1959, Journal of Physiology, 148, , figure 2. Copyright © 1959 by The Physiology Society, Cambridge University Press. Reprinted with permission.)

59 方位選擇性的例證—selective rearing
Plasticity:知覺系統因接受的刺激而產生結構與功能的改變與型塑 因為學習造成某些神經元的共同激發，進而改變cell assemblies的突觸連結(D.O. Hebb) 小貓出生後就放在黑暗環境，每天5hr在直線條筒中→假設原本應該對水平方位反應的神經元無法發展 觀察到行為上的對應變化—對於水平方位線條不敏感

60 Figure 3.31 (a) Striped tube used in Blakemore and Cooper’s (1970) selective rearing experiments. (b) Distribution of optimal orientations for 52 cells from a cat reared in an environment of horizontal stripes, on the left, and for 72 cells from a cat reared in an environment of vertical stripes, on the right. (Blakemore & Cooper, 1970). 水平線條環境 垂直線條環境

61 複雜細胞（complex cells）有運動方向選擇性（orientation vs. direction)
end-stopped cells: 除了有運動方向選擇性外，還選擇特定長度的線段或角 Hubel & Wiesel的發現顯示初級視皮質的神經元並非只因「光」刺激就反應，而是會選擇性地對於具特定空間屬性的光刺激（eg., 移動的線）作反應，這些發現為瞭解更複雜的腦知覺處理歷程奠定重要的基礎 網膜節細胞，簡單細胞，複雜細胞，end-stopped cell的特性顯示，不同視覺消息處理階段中，神經元的角色有如特徵偵測器（feature detector）--偵測到某種視覺刺激特性就開始作反應 更複雜的特徵偵測器

62

63

64 Figure 3.25 How a neuron in a monkey’s temporal lobe responds to a few stimuli. This neuron responds best to a circular disc with a thin bar. (From “Coding Visual Images of Objects in Interotemporal Cortex of the Macaque Monkey,” by K. Tanaka, H. A. Siato, Y Fukada, and M Moriya, 1991, Journal of Neurophysiology, 66, Copyright © 1981 by the American Physiological Society. Reprinted by permission.)

65 特徵偵測影響知覺的例證 運用選擇性適應（selective adaptation）技術 呈現某種神經元會作反應的刺激→該神經元作出反應
持續呈現該刺激→神經元適應後 神經元激發速率降低 如再度呈現該刺激，則神經元激發量降低 神經元激發量改變 — 神經生理反應 對於某種視覺刺激的敏感度改變 — 心理物理反應

66 選擇性適應程序 前測 測量受試者對於各種方位線條的敏感度 適應 受試者持續觀看某種方位的線條（稱為適應刺激）1-2 min
比較前後測對於不同方位線條的敏感度有無不同 選擇性適應程序 前測 測量受試者對於各種方位線條的敏感度 適應 受試者持續觀看某種方位的線條（稱為適應刺激）1-2 min 後測 測量受試者對於各種方位線條的敏感度

67 選擇性適應所採用之刺激 光柵（grating） 可以改變其方位，對比及其他屬性的刺激

68 選擇性時映所採用的依變項—對比敏感度（contrast sensitivity)
測量對比閾（contrast threshold） 運用心理物理方法，逐漸改變光柵中亮暗條的對比，直到受試可以或無法區辨出亮暗之別為止，以找出受試可以區辨亮暗的最低對比值—此為對比閾 對比敏感度=1/對比閾

69

70 選擇性適應心理物理曲線 Orientation tuning curve

71 對於大小的選擇性適應

72 神經元的size tuning curve

73 用對於大小敏感的神經元（組型）來解釋大小的選擇性適應

74 習題 #1 （選一題） （1）何謂雙眼競爭(binocular rivalry)？如何用它來研究視覺意識？你覺得這種研究方法的優點與缺點？
Crick, F.(1996). Visual perception: Rivalry and consciousness. Nature, 379, Tong, F., & Engel, S. A. (2001). Interocular rivalry revealed in the human cortical blind-spot representation. Nature, 411,

75 （2）舉例說明盲點的填補（filling-in）是什麼意思？你覺得知覺系統為什麼需要這個機制？請你試著用日常生活中的常見器物（設施）來示範盲點填補。
Ramachandran, V. S. and R. L. Gregory (1991). Perceptual filling in of artificially induced scotomas in human vision. Nature, 350,