Hierarchy of Visual Processing photons spikes NOT AOS accessory optic system Projections of the retina DTN dorsal terminal nucleus OPN LTN lateral terminal nucleus APT MTN medial terminal nucleus PPT IPm inferior pulvinar, medial region LGN lateral geniculate nucleus SC PT PGN pregeniculate nucleus PT pretectal area APT anterior pretectal nucleus SZ NOT nucleus of the optic tract SGS OPN pretectal olivary nucleus SO DTN PPT posterior pretectal nucleus IPm SC superior colliculus SC—coronal section AOS SCN suprachiasmatic nucleus SZ stratum zonale SGS stratum griseum superficiale SO stratum opticum LTN LGN MTN c P6 P5 i •K6 PGN P c 4 •K5 P3 •K4 i M2 •K3 i M1 •K2 c •K1 SCN optic tract optic chiasm optic nerve LGN—coronal section M magnocellular layer P parvocellular layer K koniocellular layer c projection from contralateral eye i projection from ipsilateral eye 166 J. H. KAAS, R. W. GUILLERY AND J. M. ALLMAN Fig. 2 The lateral geniculate nucleus of a lion in coronal and sagittal sections. The cellular discontinuity in layer A (DISC) is shown. Medial is to the left in the upper figure and rostral is to the left in the lower figure. Thionin stain. of relay laminae, probably corresponds to portion of the visual field between the the representation of the ‘line of decussa- blind spot and the zero vertical meridian. tion” or zero vertical meridian (Kaas et al., This last point allows comparison of the ’72a) the segment of the lateral geniculate visual field representations in the lion and nucleus that lies medial to the cellular dis- cat. In previous experiments (Kaas, Hoff- continuity in the lion must represent the mann, and Ladpli, unpublished observa- Responses in the retina and the LGN are more or less the same Cleland, Dubin & Levick (1971) CONTRAST SENSITIVITY IN MACAQUE L.G.N. 233 parvocellular unit, but do not differ appreciably from one another. In a sample of 105 we encountered only 6 units that showed clearly non-linear summation, and these could not be distinguished by their contrast sensitivity curves. (Magnocellular units Contrast sensitivity functions ofwere M andtypically P cells5-10 intimes macaqueCONTRASTmore sensitive SENSITIVITYLGN thanIN MACAQUEparvocellularL.G.N. units.) 233 parvocellular unit, but do not differ appreciably from one another. In a sample of CONTRAST SENSITIVITY IN MACAQUE L.G.N. 225100 A105 we encountered only 6 units that showed clearly100 non-linearB summation, and these could not be distinguished by their contrast sensitivity curves. (Magnocellular units were typically0 5-10 times more sensitive than parvocellular units.) 0 100 - A 100 0 100 A 100 B 0 0 t, 0 0~~~ .p z (A7) 0~~~~~~~~~ C 0 0~~~ 3) 10 A 10 0 0~~~~~~~~~ 0~~~~~~~~0 A 0 0~~~~~~~~0 0) 0 ~~~~~~~~~~~~~~CCP U) D 0 1 . ..... .....1*. .....l..d...... 1... .. * .. .......... 1 1 1 1 10 100 a I. A Bel a a A ago oil -1 a .... , a.,, loss .. O-'1 1 10 100 1 0.1 1 10 011 1 10 a I. A Bel a a A ago -1 a .... , a.,, loss .. Spatial frequency (cycles deg ') oil Spatial frequency (cycles deg-') 0.1 Spatial1 frequency (cycles deg-')10 Spatial01 frequency1 (cycles deg-') 10 Fig. 10. Spatial contrast sensitivity functions obtained from magnocellular units, using 100 100 F Spatialgratingsfrequencymoving at 5-2(cyclesHz. Filleddeg-')circles on the ordinates markSpatialsensitivitiesfrequencyto modulation(cycles deg-') B of a spatially uniform field. A, unit that showed linear spatial summation; B. unit that showed substantially non-linear spatial summation. ._ Fig. 10. Spatial contrast sensitivity functions obtained from magnocellular units, using at 5-2 Hz. Filled circles on the ordinates mark sensitivities to modulation 0 gratings moving C c0 10 10 of a spatiallyTheuniformsmooth curvesfield.drawnA, unitthroughthattheshowedpoints in Fig.linear10 arespatialthe best-fittingsummation;solutions B. unit that 100 showed substantiallyto eqn. (1), whichnon-linearfor these unitsspatialprovidessummation.an acceptable description of the observa- C(a tions. The fits of eqn. (1) were generally less good for magnocellular units than for so parvocellular ones. Parameter rc (the characteristic radius of the centre) is in 0 ^ 0 II U 00 0 Fig. 6C and D plotted against eccentricity for all magnocellular units whose receptive L.* field positions are known. (For one animal, in which several magnocellular units were . I . I . .. - ... ... .. I LI - The smoothstudied,curvesthedrawnrecord ofthroughreceptive fieldthepositionspointsis lost.in) AtFig.all eccentricities10 are thewherebest-fittingboth solutions 0.1 1 10 100 0.1 1 10 1100to eqn. (1), whichgroups forare representedthese unitsrc of theprovidesmagnocellularanreceptiveacceptablefield is substantiallydescriptionlargerof the observa- Spatial frequency (cycles deg-') Spatial frequency (cycles deg-') than r. of the parvocellularwere one only for units driven from ipsilateral retina. tions. The fits ofTheeqn.high contrast(1) sensitivitygenerallyof magnocellularless goodunitsforwasmagnocellularnot due to some units than for peculiarityones. Parameterof the relationship between stimulus contrast and responseradiusamplitude:of the is in 100 parvocellular rc (the characteristic centre) C F Fig. 1 1 shows this relationship for two units. The smooth curves drawn through the '100 Fig. 6C and Dpointsplottedare best-fittingagainstsolutionseccentricityto eqn. (4). forValuesallofmagnocellularK and co are much smallerunitsthanwhose receptive those for which reflects the fact that of ._ are parvocellular units,one responses magnocellular were r._ field positions known. (For animal, in which several magnocellular units (A units saturate for much lower contrasts. the between stimulus 0 :LI However, relationship C 0 0 the record of field is lost. At all eccentricities where both 0° ° ° o studied, contrast andreceptiveresponse amplitude positionswas linear over a substantial) range that included 10 10 p 0o 0 the criterion. As stimulus contrast was the of the - 0) groups are represented rc of the magnocellularprogressivelyreceptiveincreased,fieldphaseis substantially larger (U 0 response advanced by between 30 and 40 deg, possibly reflecting the operation of a (U than r. of the 'contrastparvocellulargain control'one(Shapleyonly&; Victor,for units1978). driven from ipsilateral retina. 0 c Senwitivity to temporalfrequency. For most magnocellular units contrast sensitivity UL 0 The high contrast sensitivity of magnocellular units was not due to some c) tothegratings of optimum betweenspatial frequencystimuluswas measuredcontrastat severaland temporal, , , ,,, of 1%~~~~~ ,,,,,.I,, , ,,-1% ,,,,,,,,l, ,,,,,.I, l peculiarity relationship response amplitude: 0.1 1 10 100 0.1 1 10 100Fig. 1 1 shows this relationship for two units. The smooth curves drawn through the Spatial frequency (cycles deg') Spatial frequency (cycles deg-') points are best-fitting solutions to eqn. (4). Values of K and co are much smaller than Fig. 3. Spatial contrast sensitivity of parvocellular units to sinusoidal gratings movingthose for parvocellular units, which reflects the fact that responses of magnocellular at 5-2 Hz. Filled circles on the ordinates mark the sensitivity to modulation of a spatiallyunits saturate for much lower contrasts. the between stimulus uniform field. A, C and E, type I units; B and D, type III units; F, type II unit driven However, relationship by R and G cones. contrast and response amplitude was linear over a substantial range that included the criterion. As stimulus contrast was progressively increased, the phase of the response advanced by between 30 and 40 deg, possibly reflecting the operation of a TABLE 1. Best-fitting parameters for eqn. (1) 'contrast gain control' (Shapley &; Victor, 1978). to For most units contrast Cell kc rlks r.2 r. Senwitivity temporalfrequency. magnocellularDerrington & Lenniesensitivity 1984 rs2/kc to of was measured at several A (8D) 15-03 0-015 0-89 0-072 gratings optimum spatial frequency temporal, B (14G) 9.51 0-029 0-84 0-069 C (15M) 10-74 0030 0-67 0-202 D (15A) 16-87 0-029 0-83 0-159 E (13D) 12-18 0054 037 0494 F (8E) 17-63 0-015 0-31 0020 8 PHY 357 0ARVOCELLULAR -AGNOCELLULAR A B WK WK Contrast responses of M and P cells in macaque LGN 0ARVOCELLULAR -AGNOCELLULAR AC BD A B WK WK WK WK WKPARVO !DULTParvo PARVO 0ROPORTIONOFCELLS CE Parvo DF Magno 2ESPONSEIPS WKWK WK WK L L C D WKMAGNO !DULTMAGNOMagno 0ROPORTIONOFCELLS E 2ESPONSIVITYIMPSECCONTRAST F WK WK 2ESPONSEIPS L L #ONTRAST 2ESPONSEIPS #ONTRAST 2ESPONSEIPS 2ESPONSIVITYIMPSECCONTRAST Movshon et al, 2005 -OVSHONETAL &IGURE -OVSHONETAL -OVSHONETAL &IGURE &IGURE J. Physiol. 533.2 Koniocellular cells in marmoset lateral geniculate nucleus 521 Histological processing RESULTS The position of each recorded cell was noted by reading the depth Data were obtained from 146 cells. We classified visually from the hydraulic microelectrode advance (David Kopf Model 640, responsive units as belonging to PC, KC or MC divisions of the Tujunga, CA, USA). Electrolytic lesions (6–20 µA, 6-20 s, electrode negative) were made to mark selected locations on electrode tracks. LGN if the reconstructed recording site could be clearly localised At the conclusion of recording, the animal was killed with an with respect to the laminar borders, and did not lie within 10% overdose of pentobarbitone sodium (80–150 mg kg_1, I.V.) and depth of a laminar border. This criterion reduces the sample size perfused intracardially with 0.25 l of saline (0.9 % NaCl). This was but at least partially compensates for positional uncertainties followed by 0.3 l of freshly prepared 4 % paraformaldehyde in 0.1 M which accompany the anatomical reconstruction (White et al. phosphate buffer (PB, pH 7.4) The brain was removed and placed in 1998). A total of 91 units (44 PC; 12 MC; 35 KC) met this 4% paraformaldehyde in PB for 12 h, then placed in 30% sucrose in criterion. We maximised our yield of KC cells by using PB until it sank. Coronal sections of 30 µm thickness were cut on a freezing microtome. Alternate sections were mounted onto glass relatively high impedance electrodes (10–15 MΩ) and by slides, air dried, then stained for Nissl substance.