L14. Sound Detector + Localization 2 - delay r ( ) f ( t - PDF document

9/18/2011 Linear Summation Delay-line Coincidence L14. Sound Detector + Localization 2 - delay r ( ) f ( t ) h ( ) dt Jeffress Model neural delay, Δ T September 19, 2011 BioNB4240 C. D. Hopkins coincidence 1 detectors Jeffres’ Model (1948) Barn Owl • Suppose Right & Left neural inputs from opposite sides converge.... • Delays, from neuronal conduction • Output cells (A-E) fire only if there is coincident inputs from both R and L • Delay-Line + coincidence detector produces a map of azimuth. 3 Barn Owls Owl can “localize” before leaving perch. Owl attack is precise within 2 degrees. Owl attacks sound source, not the mouse. • Conclusion: good localization in both Roger Payne established that barn vertical owls hunt in total darkness by (elevation) and horizontal (azimuth) using passive listening (Payne, 1971) * . Later work by Masakazu • How do they localize in the vertical? Konishi established the owl as an experimental model system in Neuroethology. * Payne, R. S. (1971) Acoustic location of prey by Barn owls ( Tyto alba ) 6 5 1

9/18/2011 Barn owl can locate speaker in two In barn owls, asymmetric ear placement converts IID into map of elevation dimensions y= x 2 1 0.8 0.6 0.4 Left ear 0.2 Masakazu Konishi Right ear 0 -0.2 -0.4 -0.6 -0.8 7 -1 8 0 0.5 1 Head turning does not depend on ILD Mapped feedback • Short sounds: head turn begins after the sound. • Magnitude of turn is correct for sounds of different directions. • Accuracy is 1 to 2 degrees. • Ballistic movement, no feedback needed. • Microphones in ear canals used to map difference in sound intensity 9 10 Map of ITD Discovery: “A Neural Map of Auditory Space” 11 12 BioNB424 2

9/18/2011 Individual neurons in inferior colliculus are Tuning to ITD Tuning is determined by two different “space - tuned”. Their receptive fields are oval sensory cues: ITD and ILD. shaped, tuned for sound direction. Receptive field Tuning to ILD Knudsen & Konishi (1978) 13 14 A map of auditory space Single Cells in the IC • space-tuned receptive fields • arranged in a map of auditory space • map is a “computational map” not a map of a sensory surface. 15 16 Knudsen & Konishi (1978) Top Down Case in Point We did find neurons with spatial receptive fields first in the forebrain auditory area. We could not, however, recognize any map of auditory space. Eric suggested While I was waiting for the large anechoic room that was to be that we go to the inferior colliculus. Since we could assembled in my laboratory, I worked on the owl’s visual not afford to kill an owl to make a brain atlas, we used Wulst with Jack Pettigrew. After one year of exciting research with Jack, he asked me what I was going to do with the the head of a frozen owl. We cut it with a band saw auditory system. Since I was impressed by the selectivity of and look at unstained sections to determine the visual neurons for space, I told him that I would like to map coordinates of the inferior colliculus. In the first the spatial receptive fields of auditory neurons. He multiunit recording, we saw a cluster of neurons enthusiastically endorsed this idea and paid for the necessary responding only to sound coming from a particular instruments to be built by the legendary designer and Mark Konishi direction. As I was leaving for somewhere, Eric machinist Herb Adams. The idea was to deliver sounds from continued to record further. He later reported to me various directions to plot the distribution of neuronal responses that the preferred sound direction changed in space. About this time, Eric Knudsen asked me if he could systematically as he moved the electrode. We worked join me as a postdoc. Eric and I met in a café or hamburger joint in Manhattan to discuss our project. When he asked me very hard for about three months to find a map of what I was looking for, I answered "a map of auditory space," auditory space. Eric and I celebrated this occasion Eric Knudsen Roger( ) because the owl can rapidly pin point sound sources as if it with a bottle of champagne. - Mark Konishi were using a look-up table. Eric said that he would expect a map without any reason. We were naive and ignorant because ISN Newsletter March 1999 we did not know Arnold Starr’s review * in which he pointed http://www.neurobio.arizona.edu/isn/Newsletters/isn.n out that neither spatial receptive field or maps should exist in ews.mar99.sec2.html#2 *Starr, A (1974) Fed Proc 1974 Aug;33(8):1911-4 the auditory system because space is not mapped in the inner Neurophysiological mechanisms of sound ear. localization. 18 17 3

9/18/2011 Behavioral Test of ILD Behavioral Test of ITD • Trained owls • Right ear plugged: owl • Owl presented with an turns down artificial ITD from two • Left ear plugged, turns earphones. up • Owl turns head after training to locate sounds at different azimuths • Main cue: ongoing disparity of sound times (ITD) 19 20 Hypothesis ITD Test • Behavioral test, trained owl. • Vertical axis coded by IID • ITD manipulated with stereo • Horizontal axis coded by ITD headphones. • Owl turns toward time- advanced sound. • 10 microsec sensitivity. 22 21 Cells in nucleus magnocellularis project to nucleus laminaris both ipsilaterally and contralerally. Ipsilateral axons enter dorsally; contralateral axons enter ventrally. Time pathway in the owl (blue) Carr, CE & Konishi, M 1988 1 mm NL: azimuth maps onto depth in nucleus NL analogous function to MSO in mammal 23 24 4

9/18/2011 C. Carr and M. Konishi find cells in nucleus magnocellularis that could C. Carr and M. Konishi find cells in nucleus magnocellularis that could generate a map of azimuth generate a map of azimuth 1 mm 1 mm Carr, CE & Konishi, M 1990 Carr, CE & Konishi, M 1990 NL: nucleus laminaris NL: nucleus laminaris NM: nucleus magnocellularis NM: nucleus magnocellularis 25 26 Sullivan, W. E. and Konishi, M. find neural map of ITD in owl Konishi, 2006 27 28 Map of Interaural Level Differences? VLVp (nucleus Ventralis Lemnisci Lateralis pars posterior) Takahashi, Barbarini, and Keller (1995) J. Comp. Neurol. 358:294 29 30 5

9/18/2011 Takahashi, T., C. L. Barberini, et al. (1995). "An anatomical substrate for the inhibitory gradient in the VLVp of the Owl " Journal of Comparative Neurology 359 : 294-304. VLVp VLVp injection site 31 32 Injections of tracer in the Right VLVp reveals terminals in the left dorsal VLVp Injections of tracer in the Right VLVp reveals terminals in the left dorsal VLVp The two cochlear nuclei specialized for either Time and time encoding or ampltiude encoding Amplitude • N. magnocellularis: Pathways – TIME PATHWAY – thick axons; heavy myelin Converge in – endbulb of Held the IC – adendritic cell • N. anguilaris – AMPLITUDE PATHWAY – small boutons – diffuse arbor 33 34 Vision guides plasticity Q? How does the ITD map get calibrated in development? stimulus stimulus Sound Loclization_scenes.swf Sound Loclization_scenes.swf now owl responds normally to sound, but makes 23 deg error to visual stim PNAS (1984) E. Knudsen, 1999 After 42 days, the response to sound is 35 also displaced 36 6

9/18/2011 Head size is changing The Auditory Map is Plastic 1. Earplugs cause localization errors toward plugged ear. 2. These errors are temporary : after 2-3 weeks with earplugs, owls learn correct sound location. 3. After removal of ear plugs, owls again make errors. 4. These post-removal errors are also reversible (in young owls) 5. However, adults do not compensate for ear plugs. Knudsen (1983) PNAS (1984) 37 38 Vision guides plasticity Vision is important in re-adjusting the auditory map stimulus • Errors corrected after stimulus ear plug removal (A, B) • No correction if blinders are placed on now owl responds normally to sound, owl after plug removal but makes 23 deg error to visual stim (C). • No correction if vision is displaced using prisms (D) Knudsen and Knudsen (1985) E. Knudsen, 1999 After 42 days, the response to 39 sound is also displaced 40 When recording from neurons in Vision guides plasticity Optic tectum, ITD from a neuron in the 0 degree azimuth part of the map is tuned to 0 degrees azimuth. stimulus stimulus After 8 weeks with L23 prisms, a neuron in the zero degree part of map is now tuned to 50 microsecond ITD. (the map has become displaced) now owl responds normally to sound, but makes 23 deg error to visual stim After 42 days, the response to E. Knudsen, 1999 41 42 sound is also displaced 7