Application of a commercially-manufactured Doppler-shift laser velocimeter to the measurement of basilar-membrane vibration

Abstract

A commercially-available laser Doppler-shift velocimeter has been coupled to a compound microscope equipped with ultra-long-working-distance objectives for the purpose of measuring basilar membrane vibrations in the chinchilla. The animal preparation is nearly identical to that used in our laboratory for similar measurements using the Mössbauer technique. The vibrometer head is mounted on the third tube of the microscope's trinocular head and its laser beam is focused on high-refractive-index glass microbeads (10-30 microns) previously dropped, through the perilymph of scala tympani, on the basilar membrane. For equal sampling times, overall sensitivity of the laser velocimetry system is at least one order of magnitude greater than usually attained using the Mössbauer technique. However, the most important advantage of laser-velocimetry vis-à-vis the Mössbauer technique is its linearity, which permits undistorted recording of signals over a wide velocity range. Thus, for example, we have measured basilar-membrane responses to clicks whose waveforms have dynamic ranges exceeding 60 dB.

Fig. 1

Diagram illustrating the manner of coupling between the head of the laser vibrometer and the compound microscope. The arrow heads indicate the direction of signal transmission. The dotted lines indicate lenses in the vibrometer head (3), the microscope and the microscope/vibrometer head adaptor (5). 1) Glass fiber carrying light from the laser to the vibrometer head. 2) Electrical signal from the vibrometer head to the frequency tracker. 3) Vibrometer head. 4) Bragg cell, photodiodes, prisms, etc. inside vibrometer head. 5) TV-camera adaptor: couples the vibrometer head to the third tube of the microscope trinocular head. 6) Microscope ocular: permits visual observation of target and laser-beam spot. 7) Mirrored prism: allows simultaneous visual observation of the target and laser beam transmission from the vibrometer head to the target and back. 8) Half mirror: directs incident light from the standard epi-illuminator (not shown in Figure) toward the target, while allowing the laser beam to be transmitted from the vibrometer head to the target and back. 9) Focusing knob: translates objective. 10) Post: connects the microscope to a boom stand. 11) Ultra-long-working-distance 20× objective. 12) Green acetate filter: protects experimenter’s eyes from laser light.

Fig. 2

Basilar-membrane velocity responses to 50-dB SPL tone pips with frequency 6–10 kHz. Each waveform represents averaged responses to 1024 repetitions of identical stimuli presented every 25 ms. The abscissa indicates time, in ms, after electrical stimulus onset. All responses are identically scaled. The electrical tone pip stimuli consisted of tones modulated at onset and offset by waveforms specified by cos (ø) + 1 (180° < ø c 360° at onset, 0° < ø <180° at offset). Rise/fall time was 0.82 ms. The time between application of the electrical signal to the earphone and the beginning of the offset was 3 ms.

Fig. 3

Isointensity contours for basilar membrane responses to tone pips, as a function of stimulus frequency. Tone pip characteristics were as specified in the legend for Fig. 2. Velocity amplitudes at the tone pip center frequencies (indicated in the logarithmic ordinate, with units of mm/s) were extracted by Fourier transformation of waveforms such as those of Fig. 2. The parameter is sound pressure level (re: 0.0002 dyn/cm²).

Fig. 4

Isoresponse contours (‘tuning curves’) for basilar-membrane velocity responses to tone pips recorded with the laser-vibrometer (solid lines) or using the Mössbauer technique (dashed line). The ordinate indicates the sound pressure level required, at any particular frequency, to elicit a given velocity amplitude (0.1, 0.2 or 0.4 mm/s, indicated as the parameter). For the laser vibrometry responses, sound pressure levels were interpolated logarithmically using the data of Fig. 3. For comparison, a 0.1-mm/s tuning curve is shown which is representative of results previously obtained in our laboratory using the Mössbauer technique at approximately equivalent basilar-membrane sites in normal chinchillas (Robles et al., 1986b).

Fig. 5

Illustration of the method used to extract velocity information from basilar membrane responses to clicks recorded with the Mössbauer technique (from Robles et al., 1976, with permission). Panel (a) indicates the number of gamma photons detected as a function of time from click onset. Panel (d) shows a plot of the Breit-Wigner equation, which relates photon counts to relative velocity between the Mössbauer source on the basilar membrane, and the absorber and detector. Application of the Breit-Wigner equation permits derivation of velocity curves that are full-wave rectified [panel (b)]. Full-wave rectification arises from the near-zero isomer shift of the particular source-absorber combination used. Panel (c) shows how the actual velocity waveform may be derived from the rectified waveform.

Fig. 6

Basilar membrane velocity responses to clicks, measured with laser vibrometry. Click intensity, expressed as peak pressure (re: 0.0002 dyne/cm²), is indicated for each tracing. The acoustic-click waveshape, recorded in an artificial cavity by means of a microphone with a flat frequency response, is shown at the top of the right column. Waveforms in the left column, which are identically scaled, represent averaged responses to 2048 presentations of rarefaction clicks. Electrical click onset occurred at time zero. Upward deflections indicate velocity toward Scala vestibuli. The right column displays the same data after normalization to stimulus intensity: if the responses grew linearly with click level, all waveforms in the right column would be identical. In fact, because of compressive nonlinearity, higher-level normalized responses appear smaller than responses to less intense clicks.

Fig. 7

The effect of death upon basilar-membrane responses to clicks. Each waveform is the average of 2048 responses. The left-column waveforms were recorded approximately 6 l/2 h after opening the otic capsule, while the cochlea was still in relatively good physiological condition. The waveforms on the right show basilar membrane responses several minutes after killing the chinchilla with an intravenous injection of sodium pentobarbital. Click peak pressures are indicated alongside each tracing. Note that response amplitudes in the live preparation change less with stimulus intensity than those in the dead chinchilla.

Fig. 8

Frequency spectra of basilar-membrane velocity responses to clicks in a relatively healthy cochlea and in the same cochlea, post mortem. The frequency spectra were obtained by Fourier transformation of the time waveforms shown in Fig. 7. Spectra are shown for responses at 75 and 95 dB SPL for each of the healthy and dead preparations. Responses are much more sensitive and frequency selective in the five cochlea. Note that the frequency spectra, expressed in units of velocity per unit pressure, are strongly dependent on stimulus level in the live cochlea (and hence strongly nonlinear), while almost perfectly linear in the dead chinchilla.